Author Topic: 100:1 probe for measuring ripple in a tube amp power supply  (Read 11243 times)

0 Members and 1 Guest are viewing this topic.

Offline Hydron

  • Frequent Contributor
  • **
  • Posts: 978
  • Country: gb
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #25 on: January 29, 2019, 07:07:21 am »
The OP is 100% correct to be thinking of the 400V rating - coupling mode DOES matter and AC coupling will negate any probe scaling factor for DC input for most probe/frontend designs. He/she understands a subtle gotcha in the way the AC coupling mode works and is asking a reasonable question rather than trying to start an argument.

I'd also agree with the suggestion above to use DC coupling and an external capacitor at the tip however.
 
The following users thanked this post: schmitt trigger, Andreax1985

Offline Andreax1985Topic starter

  • Regular Contributor
  • *
  • Posts: 64
  • Country: it
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #26 on: January 29, 2019, 07:21:37 am »
When you're using a 1x probe, it's closer than I'd cut it but it's within spec.  When you're using a 10x probe, sure, cause that's only 32V on the actual frontend of the scope.  Just make sure your probe is rated well enough.

Please notice that I'am referring to AC coupling mode, so the blocking cap is charged at full 320V even when using attenuating probes. So, since I'm not sure of the blocking cap's rating, I can only infer it from the scope input rating (400V pk). BUT: is 400V pk enough for 320V DC?

Input rating is NOT 400V pk.
It is:

1 MΩ: ≤400 Vpk(DC + Peak AC <=10 kHz)

If this is difficult to understand and you are working with over 300V voltages I think it is not oscilloscope at all what you need first.
First you need some basic fundamentals study book. Yes I know reading is boring but... 
but also not alone reading, all need also understand.  But it is good you ask. This is wise.

But then you do not listen, you repeat and repeat same question without giving any more real facts about signals what you need measure. If you only tell 320V DC and you need measure ripple how you think it can answer anymore than example +320V +  +80V is +400V  (400Vpk)  also -320V + -80V is  -400V (400Vpk) You can even alternate between these as long as you stay below 10kHz (sine). (so <10kHz sine <800Vpk-pk (note small but important difference between pk and pk-pk) is accepted and with specification limits)
Without more details I think many peoples can repeat these answers infinitely.



You can connect 400V DC to scope input. But remember also what is DC.  When you connect it, it is not DC at all. How long time after step you can call it DC.  Think carefully this 10kHz. When you switch from 0V to 400V it depends this edge what is maximum frequency what there exist. So you can "slowly" (yes 10kHz is slow) rise it from 0V to 400V..  to +400 or  to -400V (ref GND)  and same when you change it from this voltage back to 0. You need stay below 10kHz if you are workin there near max limits.

(btw, pure DC is possible only in human imagination. In practice there is not DC at all but yes, perhaps 1pHz can think in practice as "DC" perhaps even nHz or uHz).   ;)

But, can you explain more what you really need measure. You have tild 320V (dc) and then you want measure some ripple in this "dc".  What kind of ripple. What is this ripple level range and max freq.

Btw, as far as I know Siglent use 500V rated capacitors in front end these positions what need it.
(it can also see in this old SDS1202X-E "compensation issue" case)

But, can you explain more what you really need measure. You have tild 320V (dc) and then you want measure some ripple in this "dc".  What kind of ripple. What is this ripple level range and max freq.

If you afraid this internal AC coupling capacitor you can of course use external capacitor and then oscilloscope DC coupled. This give also one more advantage. You can select  what is low frequency corner what you need instead of Siglent very low corner freq around 1.2-1.3Hz. (more low than many other scopes and if your DC have low f fluctuations then ripple measurement may be extremely difficult if you need look low levels.) Is is possible or useful to use scope DC coupled and external coupling capacitor?

What is this ripple (over this 320V) what you want measure. Ripple pk-pk level and something about freq range and about what V/div,  mV/div range you need for ripple measurement?

Thanks for the 500V bit of info on the coupling capacitor rating. As for the rest I can't be precise because I can't know before I probe the circuit. But since what I want to probe is a classic power supply stage in a tube (audio) amp I think ripple will be very low compared to the Dc signal (so no worries about it's magnitude). Also, I expect it to be at 100hz frequency as typical in this kind of applications. So I have no worries related to its frequency either.
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #27 on: January 29, 2019, 07:44:10 am »
Thanks for the 500V bit of info on the coupling capacitor rating. As for the rest I can't be precise because I can't know before I probe the circuit. But since what I want to probe is a classic power supply stage in a tube (audio) amp I think ripple will be very low compared to the Dc signal (so no worries about it's magnitude). Also, I expect it to be at 100hz frequency as typical in this kind of applications. So I have no worries related to its frequency either.

A 10:1 300Vrms CAT II probe is enough for your measurement, including quite a good deal of margin. This is because 300Vrms equals to 420Vpk, and CAT II includes headroom for overvoltage. And your DC source is of limited energy (discharge of the filter cap, gives you a loud bang when shorted, but no violent explosion).
I agree to the others, a 10:1 probe is safer than a 1:1 probe here (for measuring at 320V DC) and will give you enough display resolution as one expects the ripple voltage in the "some volts" range here. The internal resistance of the 10:1 probe will do a good job of protecting the scope's input.

The general strategy for measuring ripple on a supply rail would be:
Set the scope to a high enough voltage / div scale, to cover the DC voltage, set coupling to DC.
Connect the probe, turn on the DUT (in any order you like)
Watch the signal for plausibility, check it's not exceeding the range.
As long as you can display the signal in DC coupling, one can assume it's safe to switch to AC coupling.
Change to AC coupling and increase sensitivity (less volts / div) until you can watch the ripple.
Disconnect and discharge the AC coupling capacitor by setting the scope to DC coupling again.

This procedure takes care of safely charging / discharging the scope's AC coupling capacitor. Usually these steps are skipped, because most general purpose scope's inputs are robust enough to not require them. Especially when you're using a probe (even the 1:1 ones should have some internal losses to dampen the charge / discharge currents) and operating at typical supply voltages up to 20V. For your 320V, I'd recommend to be more careful in general, since a charged capacitor at 320V bites you and your electronics. A charged AC coupling capacitor (from measuring B+) can damage small signal semiconductors if you move the probe from B+ to a small signal node without discharging the AC coupling capacitor, even when using a 10:1 probe.
Safety devices hinder evolution
 
The following users thanked this post: Andreax1985

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1773
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #28 on: January 29, 2019, 07:44:32 am »
Hi, I wish to measure b+ ripple in a tube amp. Since the b+ rails are at 350V, I was considering buying a 100:1 passive probe for my scope (rated for 400V input). Nevertheless, here is my doubt: I'll need to use the scope in AC coupling mode to clearly see the ripple. This means that I cannot safely use a cheap 100Mohm 100:1 probe. I'm on a budget and I can't afford a differential probe. Could you suggest me a suitable 100:1 probe which does not give issues when used in AC coupling mode?

This is how I did it (to measure ripple in an 500V HV power supply):

- a took a *big* foil cap (10uF@1000V) and connected one end to the PSU and grounded the other end. At the grounded side
it looks like a resistor (e.g., 10kOhm) and a pair of antiparallel diodes (1N4007) that are bridged by a switch, now closed.
- I connected the load to the power supply
- I power up the PSU. The cap now charges up thru the closed ground switch
- I open the ground switch. AC now appears across the 10kOhm resistor and can be measured. Lower bandwidth limit is ca. 10Hz.
- Performing measurements
- Afterwards, close the ground switch again
- Disconnect scope
- Power down the PSU
- After coupling cap is empty, disconnect it
- Done
 
The following users thanked this post: Andreax1985

Offline trobbins

  • Frequent Contributor
  • **
  • Posts: 755
  • Country: au
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #29 on: January 29, 2019, 07:44:43 am »
The B+ ripple in a valve amp may well get to 5-10Vrms for a capacitor input filter, especially with a valve diode.  Be very careful to check if the power supply uses a choke input filter, as the diode output voltage that connects to the choke will be a large DC+AC level.  Most valve amp people just get a nice high voltage cap of about 10nF and place it between the B+ and a 10:1 probe, and use DC coupling mode to ensure the probe and scope parts are all ground referenced.  Clipping the cap/probe to the test point, and then energising the amp, and then de-energising, allows the interface cap to charge up and down more sedately and safely.  A 10:1 probe presents 10Meg loading, which is usually high enough not to noticeably alter voltage rail levels in the preamp sections of an amp (whereas 1Meg loading can have a noticeable influence).
« Last Edit: January 29, 2019, 07:47:03 am by trobbins »
 
The following users thanked this post: Old Printer, Andreax1985

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6453
  • Country: hr
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #30 on: January 29, 2019, 09:03:12 am »
Choose right tool for the job.
You cannot rely on AC decoupling in the scope to do this without blowing up.

You should buy a book "Building Valve Amplifiers, Second Edition" by Morgan Jones.
Books are our friends.

In attachment excerpt from the book about your particular task.

Regards,
Sinisa
 
The following users thanked this post: Old Printer, Andreax1985

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #31 on: January 29, 2019, 11:06:50 am »
When you're using a 1x probe, it's closer than I'd cut it but it's within spec.  When you're using a 10x probe, sure, cause that's only 32V on the actual frontend of the scope.  Just make sure your probe is rated well enough.

Please notice that I'am referring to AC coupling mode, so the blocking cap is charged at full 320V even when using attenuating probes. So, since I'm not sure of the blocking cap's rating, I can only infer it from the scope input rating (400V pk). BUT: is 400V pk enough for 320V DC?

Input rating is NOT 400V pk.
It is:

1 MΩ: ≤400 Vpk(DC + Peak AC <=10 kHz)

If this is difficult to understand and you are working with over 300V voltages I think it is not oscilloscope at all what you need first.
First you need some basic fundamentals study book. Yes I know reading is boring but... 
but also not alone reading, all need also understand.  But it is good you ask. This is wise.

But then you do not listen, you repeat and repeat same question without giving any more real facts about signals what you need measure. If you only tell 320V DC and you need measure ripple how you think it can answer anymore than example +320V +  +80V is +400V  (400Vpk)  also -320V + -80V is  -400V (400Vpk) You can even alternate between these as long as you stay below 10kHz (sine). (so <10kHz sine <800Vpk-pk (note small but important difference between pk and pk-pk) is accepted and with specification limits)
Without more details I think many peoples can repeat these answers infinitely.



You can connect 400V DC to scope input. But remember also what is DC.  When you connect it, it is not DC at all. How long time after step you can call it DC.  Think carefully this 10kHz. When you switch from 0V to 400V it depends this edge what is maximum frequency what there exist. So you can "slowly" (yes 10kHz is slow) rise it from 0V to 400V..  to +400 or  to -400V (ref GND)  and same when you change it from this voltage back to 0. You need stay below 10kHz if you are workin there near max limits.

(btw, pure DC is possible only in human imagination. In practice there is not DC at all but yes, perhaps 1pHz can think in practice as "DC" perhaps even nHz or uHz).   ;)

But, can you explain more what you really need measure. You have tild 320V (dc) and then you want measure some ripple in this "dc".  What kind of ripple. What is this ripple level range and max freq.

Btw, as far as I know Siglent use 500V rated capacitors in front end these positions what need it.
(it can also see in this old SDS1202X-E "compensation issue" case)

But, can you explain more what you really need measure. You have tild 320V (dc) and then you want measure some ripple in this "dc".  What kind of ripple. What is this ripple level range and max freq.

If you afraid this internal AC coupling capacitor you can of course use external capacitor and then oscilloscope DC coupled. This give also one more advantage. You can select  what is low frequency corner what you need instead of Siglent very low corner freq around 1.2-1.3Hz. (more low than many other scopes and if your DC have low f fluctuations then ripple measurement may be extremely difficult if you need look low levels.) Is is possible or useful to use scope DC coupled and external coupling capacitor?

What is this ripple (over this 320V) what you want measure. Ripple pk-pk level and something about freq range and about what V/div,  mV/div range you need for ripple measurement?

Thanks for the 500V bit of info on the coupling capacitor rating. As for the rest I can't be precise because I can't know before I probe the circuit. But since what I want to probe is a classic power supply stage in a tube (audio) amp I think ripple will be very low compared to the Dc signal (so no worries about it's magnitude). Also, I expect it to be at 100hz frequency as typical in this kind of applications. So I have no worries related to its frequency either.

You can use PP510 or PP215 siglent probe.
Set oscilloscope to DC cxoupling.
Set probe to 10x
Set scope for 10x probe
Connect 2.2nF capacitor rated least 500V (or more V)  to DUT test point and other end of capacitor to probe tip. BW low end corner frequency is bit under 10Hz now. (if calculate it supoer simply using RC formula) (many scopes AC coupling is also in this ballpark. But Siglent 1000X-E AC coupling corner freq is around 1.2Hz and this is quite low. Due to this if your 320V DC have some low freq instability is make very difficult to measure ripple because if you use low vertical setting like 5 - 20mV/div your trace difficult to observe because it is jumping out from display.

And if you look AC line frequency related ripple, as some told previously. Using trigger source "AC line" is useful.

If you want use 1x probe then these Siglent probes are not suitable. But under 300V same with 220nF capacitor result is same and then can use down to 0.5mV/div with 20MHz BW on..  If really need go so high sensitivity. 20MHz BW can turn on always when use 1x probe because probe itself have more low freq response buit 20MHz BW reduce scope front end wide BW noise.

Of course also you can - if really want - use scope own AC coupling with voltages up to 400Vpk.  Scope data sheet limits are still there (400Vpk  400V dc. ). If there is different limit for AC coupling it need also read in data sheet. There is no, so if you connect 320V DC to input and it is AC coupled and it blows ask new main board or repair from Siglent because data sheet (promise) tell it can do.

I want also sidenote one thing about probes. I have seen many peoples who do not look other than probe "nameplate" rating. There is other thing but because there all users do not read these "small things".
There is same kind of thing with what ever brand probes, including Keysight, Tektronix etc.. some differencies in numbers but same principle. Max voltage is highly dependent of frequency (mostly)
Example probe PP215 (200MHz probe what come with 1000X-E 200MHz models.
There nicely read in probe. 10x 600VDC, Pk, AC.
BUT
It is only for up to roughly around 50kHz
100kHz 200V
1MHz it is only 50V and 100MHz only around 25V
« Last Edit: January 29, 2019, 11:13:12 am by rf-loop »
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #32 on: January 29, 2019, 08:42:42 pm »
As pointed out several times, common x10 and x100 probes do not attenuate the DC component when the oscilloscope input is set to AC coupled mode and this is a serious hazard to the oscilloscope.

Better high voltage probes including x10 models include a shunt resistance but unfortunately they are expensive and rare.  They can be identified from their input resistance specification which will be lower so a x10 probe might only be 1 megohm and a x100 probe might be 10 megohms.

A better and safer way to make this measurement without a suitable probe is to use an external AC coupling capacitor and shunt resistance to remove the high voltage DC component with a standard x1 or x10 probe.  The higher input resistance of a x10 probe will yield a lower cutoff frequency for a given coupling capacitance which may be important in line frequency applications for accurate results.
 
The following users thanked this post: Andreax1985

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #33 on: January 30, 2019, 02:39:11 am »
Most of the time, if you had an old analog 'scope,you could pretty much "wing it" as they don't seem to be as fragile as DSOs.
Additionally, the really old Tektronix probes (the big,clunky ones) were designed for use with tube equipment, hence their voltage ratings are better, removing that source of worry.

As I mentioned before, some Oscilloscopes have a DC offset function, which allows you to view small ac signals superimposed upon large DC voltages, whilst in the "DC coupled" mode.

You can also do this with your 'scope in "A-B" mode, using two channels, with one being fed a DC voltage which becomes the "DC offset".
The problem here, is finding an adjustable high voltage supply which is beyond reproach as far as hum, etc are concerned---you don't want to be seeing imperfections in your "offset voltage"!

A way to use your "ac coupled" mode without,any threat to your coupling capacitor, is to make up a voltage divider, with a fairly high total resistance, but with the bottom, & smallest, resistance having a voltage drop across it well within the 'scope ratings.
Hang the voltage divider across the HT line, & look at the hum level on that lower voltage to your heart's content.

Obviously, the total resistance of the divider must be high enough not to draw too high a current, otherwise, your results will be in error.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #34 on: January 30, 2019, 04:45:27 am »
As I mentioned before, some Oscilloscopes have a DC offset function, which allows you to view small ac signals superimposed upon large DC voltages, whilst in the "DC coupled" mode.

A Tektronix 7A13 differential comparator has an offset range of +/-500 volts (and maximum input of +/-500 volts) at a 0.1V/div sensitivity.  Or a x100 probe could be used with DC coupling and 1mV/div sensitivity to yield 0.1V/div sensitivity with a +/-1000 volt offset range although it would be pretty noisy unless the 5MHz bandwidth limit is used.

The same 7A13 has AC input coupling capacitors good to 500 volts though which is higher than common oscilloscope inputs so it could also be used directly in AC coupled mode for 1mV/div sensitivity (noisy again) with a suitable x1 probe if one could be found.
 

Offline Andreax1985Topic starter

  • Regular Contributor
  • *
  • Posts: 64
  • Country: it
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #35 on: January 30, 2019, 09:23:06 am »
All in all, I think that the safest solution is to use DC coupling mode with an external AC coupling capacitor (even if it sickens me not to being able to safely use the AC coupling mode in my scope, if they feel that I cannot use full input in AC mode they should have specified it in the datasheet). Now, for computing the minimum value of this capacitor I'm using the formula f = 1/(2*pi*R*C) where f is the desired -3db frequency of the high pass filter (let's say 2Hz), and R is the total resistance (let's say 10Mohm when using 10x probes). This yields C = 10nF. If I wanted to cut frequencies below 15Hz I'd choose C = 1nF. Is this correct?
« Last Edit: January 30, 2019, 10:00:57 am by Andreax1985 »
 

Offline trobbins

  • Frequent Contributor
  • **
  • Posts: 755
  • Country: au
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #36 on: January 30, 2019, 09:52:33 am »
To minimise amplitude differences between mains and 2nd harmonic, perhaps aim for one decade below mains frequency.
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #37 on: January 30, 2019, 10:05:24 am »
All in all, I think that the safest solution is to use DC coupling mode with an external AC coupling capacitor (even if it sickens me not to being able to safely use the AC coupling mode in my scope, if they feel that I cannot use full input in AC mode they should have specified it in the datasheet). Now, for computing the minimum value of this capacitor I'm using the formula f = 1/(2*pi*R*C) where f is the desired -3db frequency of the high pass filter (let's say 2Hz), and R is the total resistance (let's say 10Mohm when using 10x probes). This yields C = 10nF. If I wish to cut frequencies below 15Hz I'd choose C = 1nF. Is this correct?

Siglent internal AC coupling is designed so that it meets front panel information. 400Vpk. You can connect 400V DC to oscilloscope input using internal coupling DC or AC.

How ever it is, it is still good practice to use external coupling capacitor in Your case what you have told also for reduce scope internal parts stress.

Previously I have also told about external coupling capacitor and freq response.
Yes your calculus is ok. (Roughly because it is complex network and not only ideal R and C. In practice still well enough accurate.)



Also this can use
You can give known/wanted values and it calculate unknown.
http://www.learningaboutelectronics.com/Articles/High-pass-filter-calculator.php

As told previously Siglent this model internal AC coupling have quite low corner frequency. Around 1.2Hz what is low. Perhaps too low for many applications. (if example testing ripple this "dc" under test may have too much drift/fluctuation so with sensitive V/div settings it is difficult to observe due to large vertical movements. So, with external AC coupling it is easy to select suitable corner frequency.

Btw, when use input coupling DC with also external DC coupling,  it is good to note that you can select vertical mode for fixed offset voltage or for fixed vertical position and this last one is handy in many cases when looking some small signal riding over DC. Now when adjust V/div range it keep trace position. This feature can use even for lot of more high resolution DC level (indirect) measurements using offset because offset DAC is lot of more than 8bit. 10x probe, 5mV/div and offset max 20V and offset resolution 100uV.
(and same for horizontal, there can select fixed delay time mode or fixed position mode (example for keep trigger position other than center of screen. Example fixed position in nearly left side of screen and it stay fixed independent of t/div adjusting)
« Last Edit: January 30, 2019, 10:10:16 am by rf-loop »
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline trobbins

  • Frequent Contributor
  • **
  • Posts: 755
  • Country: au
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #38 on: January 30, 2019, 10:42:58 am »
Also note that valve amp turn on can cause a much higher voltage than measured at idle,  as diodes conduct before typical amp valves.   A scope interface should aim to cope with at least the peak of the power transformer hv secondary ac voltage, plus margin.
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6453
  • Country: hr
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #39 on: January 30, 2019, 11:56:22 am »
I would not use anything that is not safe up to 1000V on valve amplifiers..
If you include peak voltages, lifted grounds, grid voltage variations, wrongly positioned voltage selectors, and you need some reserve..
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #40 on: January 30, 2019, 12:17:46 pm »
All in all, I think that the safest solution is to use DC coupling mode with an external AC coupling capacitor (even if it sickens me not to being able to safely use the AC coupling mode in my scope, if they feel that I cannot use full input in AC mode they should have specified it in the datasheet).

400 volts is the most common design specification because it allows work on 340 volt DC produced from rectified 240 volt AC power like would be found in an off-line switching power supply.

Quote
Now, for computing the minimum value of this capacitor I'm using the formula f = 1/(2*pi*R*C) where f is the desired -3db frequency of the high pass filter (let's say 2Hz), and R is the total resistance (let's say 10Mohm when using 10x probes). This yields C = 10nF. If I wanted to cut frequencies below 15Hz I'd choose C = 1nF. Is this correct?

10nF for 1.6Hz and 1nF for 16Hz for a x10 probe, yes, I get the same result.  Using a x10 probe with its 10 megohm input resistance makes the external AC coupling capacitor much more manageable.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #41 on: January 30, 2019, 12:24:14 pm »
Also note that valve amp turn on can cause a much higher voltage than measured at idle,  as diodes conduct before typical amp valves.   A scope interface should aim to cope with at least the peak of the power transformer hv secondary ac voltage, plus margin.
I think a better word would be a "somewhat" higher voltage.

Traditionally, valve amplifiers used thermionic rectifiers, which heated up at a similar rate to the other valves.
They also used LC filtering which tended to limit any excessive voltage at the output of the supply.
Silicon diodes & RC filtering may not be as good at this.

The easy way to avoid such higher voltages is to turn the amp "on" first, but that means you have to connect the probes whilst the HT line is powered up, which would not be advisable for the beginner.

It is still an undeniable fact that people used Oscilloscopes to probe valve device HT lines for decades with no adverse effect upon either the instrument, or the operator.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #42 on: January 30, 2019, 12:46:03 pm »
 author=2N3055 link=topic=166216.msg2162602#msg2162602 date=1548849382]
I would not use anything that is not safe up to 1000V on valve amplifiers..
If you include peak voltages, lifted grounds, grid voltage variations, wrongly positioned voltage selectors, and you need some reserve..
[/quote]

I have never seen a "1000v rated" Oscilloscope probe used over the many years in which I, & many others,  worked on valve equipment of all kinds.
The old Tektronix probes were 500v rated, which was quite adequate.

Any tech worth their salt would check for "wrongly positioned voltage selectors", which would be quite likely to have caused damage to the amplifier anyway.

"Peak voltages".
Do you mean resonant peaks?
Or just the normal peak voltage of 1.414 x RMS?

"Grid voltage variations?" They may cause changes in the current draw of tubes, but I can't see a situation where that would increase the voltage above that of the HT supply.

"Lifted grounds? "In what context?
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6453
  • Country: hr
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #43 on: January 30, 2019, 03:17:09 pm »

I have never seen a "1000v rated" Oscilloscope probe used over the many years in which I, & many others,  worked on valve equipment of all kinds.
The old Tektronix probes were 500v rated, which was quite adequate.

Any tech worth their salt would check for "wrongly positioned voltage selectors", which would be quite likely to have caused damage to the amplifier anyway.

"Peak voltages".
Do you mean resonant peaks?
Or just the normal peak voltage of 1.414 x RMS?

"Grid voltage variations?" They may cause changes in the current draw of tubes, but I can't see a situation where that would increase the voltage above that of the HT supply.

"Lifted grounds? "In what context?


There are many 1000V +  rated probes, passive and active. Passive ones will be 100:1.
Any tech worth of salt also sometimes makes mistakes. Old amps used to have several voltages 110/120/130--210/220/230/240V. Selector on 210V plugged into today's 230V distribution in Croatia  (Un +10 % / -15% = max.: 253 V, min: 199,5 V) would not mean instant explosion usually, but would mean that voltage on the secondary would rise 20%, so 400V would become 480V.
I'w seen valve power amplifiers with 500V+ inside... 20% on top of that is a lot.
Both resonances, fast unloading of large inductors in power filters, and yes P-P as opposed to RMS voltages.
Lifted grounds as in parts of circuits left floating, and having weird voltages depending where you measure from...

Thank you for sharing your experience.
You obviously know much more about topic than me so you can avoid all pitfalls. I like to be sure that my lack of such wholesome knowledge doesn't end up killing expensive piece of equipment.
That's all. And also, "old school" CRT scopes were much more robust than those made in last, let's say, 20 years. Partially because they were made to service such equipment, and partially because of high quality and over-engineering.  You could use it without worry.

My (very expensive) MSO3000T says 135V RMS on front panel, next to input BNC.  In manual it says: "CAUTION When measuring voltages over 30 V, use a 10:1 probe."


Not meant to be used on tube amps, at least not without high voltage active probe...

Regards,
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 19281
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #44 on: January 30, 2019, 03:36:30 pm »

I have never seen a "1000v rated" Oscilloscope probe used over the many years in which I, & many others,  worked on valve equipment of all kinds.
The old Tektronix probes were 500v rated, which was quite adequate.

Any tech worth their salt would check for "wrongly positioned voltage selectors", which would be quite likely to have caused damage to the amplifier anyway.

"Peak voltages".
Do you mean resonant peaks?
Or just the normal peak voltage of 1.414 x RMS?

"Grid voltage variations?" They may cause changes in the current draw of tubes, but I can't see a situation where that would increase the voltage above that of the HT supply.

"Lifted grounds? "In what context?

There are many 1000V +  rated probes, passive and active. Passive ones will be 100:1.

... except those that are 1000:1 passive probes, e.g.
http://w140.com/tekwiki/wiki/P6013
http://w140.com/tekwiki/wiki/P6015

Note that those do have the full potential divider inside the probe (i.e. don't rely on the scope's 1Mohm input resistance), so they can safely be used with the scope internally AC coupled.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6453
  • Country: hr
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #45 on: January 30, 2019, 03:49:30 pm »

... except those that are 1000:1 passive probes, e.g.
http://w140.com/tekwiki/wiki/P6013
http://w140.com/tekwiki/wiki/P6015

Note that those do have the full potential divider inside the probe (i.e. don't rely on the scope's 1Mohm input resistance), so they can safely be used with the scope internally AC coupled.

Sorry, I meant to relay that they won't be 10:1 or 1:1 probes..

But we are digressing.

OP asked how to measure ripple on top for 300-500V power supply. Way to do it (with common scopes that are widespread today) is to have a probing solution that will decouple DC in a safe manner and won't attenuate AC. Also you want to highpass filter it so anything less than 10-20 Hz doesn't mess up measurements.

In a book I quoted there is a good chapter on topic.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #46 on: January 30, 2019, 03:58:10 pm »
And except those old proper Tektronix x100 passive probes:

http://w140.com/tekwiki/wiki/P6007
http://w140.com/tekwiki/wiki/P6009

Which are safe to use with any coupling up to 1.5 kilovolts because they also include an internal divider.
 

Offline cvanc

  • Frequent Contributor
  • **
  • Posts: 672
  • Country: us
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #47 on: January 30, 2019, 05:30:29 pm »
What am I missing here?  Wouldn't measuring it with a meter be both easier and more repeatable?
 
The following users thanked this post: JOFlaherty

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 4064
  • Country: fi
  • Born in Finland with DLL21 in hand
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #48 on: January 30, 2019, 05:50:23 pm »
What am I missing here?  Wouldn't measuring it with a meter be both easier and more repeatable?

You have missed OP's first message in this thread. ;)
I drive a LEC (low el. consumption) BEV car. Smoke exhaust pipes - go to museum. In Finland quite all electric power is made using nuclear, wind, solar and water.

Wises must compel the mad barbarians to stop their crimes against humanity. Where have the wises gone?
 

Offline eliocor

  • Supporter
  • ****
  • Posts: 519
  • Country: it
    • rhodiatoce
Re: 100:1 probe for measuring ripple in a tube amp power supply
« Reply #49 on: January 31, 2019, 12:53:39 am »
Take a look at the whole application note AN118:
https://www.analog.com/media/en/technical-documentation/application-notes/AN118fb.pdf
lots of suggestions on how to measure ripple voltages and noise on HV power supplies.
 
The following users thanked this post: WhichEnt2


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf