Author Topic: Trimpot rating when in place of two resistors??  (Read 3312 times)

0 Members and 1 Guest are viewing this topic.

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Trimpot rating when in place of two resistors??
« on: February 12, 2019, 12:35:19 pm »
Hi All,

I have a schematic for my amplifier circuit board that shows two resistors on either side of an NPN transistor, 1 750ohm .25w and one 4750ohm .15w. However, on the actual circuit the resistors have substituted for a 10kohm trimpot for bias adjustment. I have been advised to replace the old carbon trimpot for a modern cermet one for good measure.

However, as the original resistors were rated at .25w each, do I need a .5w trimpot or .25w?

Thanks.
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: Trimpot rating when in place of two resistors??
« Reply #1 on: February 12, 2019, 12:54:26 pm »
It's going to depend on the voltage across the pot.

Can you show us the schematic - and point out the relevant location?
 

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #2 on: February 12, 2019, 01:03:01 pm »
It's going to depend on the voltage across the pot.

Can you show us the schematic - and point out the relevant location?

See attached.

The trimpot is used in place of R11 & R12 to the base of Q6, both resistors are .25w.
 

Offline alex.martinez

  • Contributor
  • Posts: 44
  • Country: es
  • Happiness is fixing electronics by hitting them.
Re: Trimpot rating when in place of two resistors??
« Reply #3 on: February 12, 2019, 01:37:42 pm »
It's going to depend on the voltage across the pot.

Can you show us the schematic - and point out the relevant location?

See attached.

The trimpot is used in place of R11 & R12 to the base of Q6, both resistors are .25w.

You should be good with a 1/4 W resistor. It's for bias, so don't expect a huge current through it.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: Trimpot rating when in place of two resistors??
« Reply #4 on: February 12, 2019, 02:06:08 pm »
R11+R12 should *NOT* be replaced by a trimpot as that prevents the bias limiting safety circuit (D3, R13) working properly if you turn the trimpot too far.

Instead R11 should be replaced by the trimpot, wired as a variable resistor, keeping R12 as specified on the schematic.   

Provided you refit R12, the trimpot's dissipation will be negligible as there should never be more than 4V across the Vbe multiplier.
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: Trimpot rating when in place of two resistors??
« Reply #5 on: February 12, 2019, 04:40:53 pm »
Last piece of advice: use a trimpot with a good temperature coefficient. Avoid like the plague the cheap carbon film trimpots.

Event then, the trimpot's TC will most likely be larger than most good metal film resistors.
 

Offline MrAl

  • Super Contributor
  • ***
  • Posts: 1666
Re: Trimpot rating when in place of two resistors??
« Reply #6 on: February 12, 2019, 05:08:32 pm »
Hi,

If you have two resistors rated 1 watt each and connect them in series then the total wattage rating is 2 watts.
Now if you use the center tap of those two, each resistor is again 1 watt.
Therefore when you tap a resistor in the center that is 2 watts, each half is 1 watt.
Therefore when you tap a resistor 1/10 from one end that is 2 watts, that end is 0.2 watts and the rest is 1.8 watts.
So you see how this works.
For a 10k pot that is 1/2 watt adjusted to the center position, each half section is 1/4 watt.
For a 10k pot that is 1/2 watt adjusted to 1/4 from one end, that end is 1/8 watt and the rest is 3/8 watt.
This is because the temperature rise is proportional to an exponential power of P/A where P is power and A is surface area.  If you cut the surface area A by 1/2 then if you cut the power P also by 1/2 then the temperature rise (for both sections) is the same as it is for the entire section with the full power.
« Last Edit: February 12, 2019, 05:14:57 pm by MrAl »
 

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #7 on: February 12, 2019, 05:20:02 pm »
Thank you everyone, very helpful and I will indeed follow the advice.

I have gone with a couple of 10k cermet trimpots with a TCR of 100ppm.

One last question, in regards to the original schematic, is the attached image the correct way to reintegrate R12 and the new trimpot??

Thanks again.
 

Online Gyro

  • Super Contributor
  • ***
  • Posts: 10173
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #8 on: February 12, 2019, 05:28:59 pm »
R11+R12 should *NOT* be replaced by a trimpot as that prevents the bias limiting safety circuit (D3, R13) working properly if you turn the trimpot too far.

Instead R11 should be replaced by the trimpot, wired as a variable resistor, keeping R12 as specified on the schematic.   

Provided you refit R12, the trimpot's dissipation will be negligible as there should never be more than 4V across the Vbe multiplier.

Personally, I would keep both R11 and R12 and put a much lower value trimpot (eg. 100R) at their junction. Replacing either one with a trimpot is going to make fine quiescent current adjustment pretty much impossible. It would also minimise the adverse effect of the TC of the trimpot.
Best Regards, Chris
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Trimpot rating when in place of two resistors??
« Reply #9 on: February 12, 2019, 06:46:57 pm »
R11+R12 should *NOT* be replaced by a trimpot as that prevents the bias limiting safety circuit (D3, R13) working properly if you turn the trimpot too far.

Instead R11 should be replaced by the trimpot, wired as a variable resistor, keeping R12 as specified on the schematic.   

Provided you refit R12, the trimpot's dissipation will be negligible as there should never be more than 4V across the Vbe multiplier.

Personally, I would keep both R11 and R12 and put a much lower value trimpot (eg. 100R) at their junction. Replacing either one with a trimpot is going to make fine quiescent current adjustment pretty much impossible. It would also minimise the adverse effect of the TC of the trimpot.
I agree. It should look like this.

Note to the original poster: please do a little image processing, rather than posting monstrous image files. I've converted the file to monochrome, cropped and scaled it down. Not only is the file a under a 100th of the size, but it's easier to read.
« Last Edit: February 12, 2019, 06:48:37 pm by Zero999 »
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: Trimpot rating when in place of two resistors??
« Reply #10 on: February 12, 2019, 10:04:56 pm »
@Zero999: Unfortunately 'R10' is already on the original schematic, its the emitter resistor of the current sink Q4.

@All,
Please note the PMD16K100 and PMD17K100 output transistors are complimentary Darlingtons.  I didn't realise that originally, so my estimate for the max bias voltage across the Vbe multiplier Q6 was wrong.   Due to the complexity introduced by the bias limiter (D3, R13), and the varying Vbe drops with quiesent current, I didn't fancy trying to solve the bias cirucit analytically so I have knocked up a quick rough sim of it in LTspice with only one pair of output transistors and default transistor models with only HFE tweaked to get the output transistors in the right ballpark for power Darlingtons of the specified types.  Component designators correspond to the original schematic.

It shows a bias point of approx 4V across Q6 and 6.6mA quiescent current per pair of output transistors. Edit: I got the emitter resistr values wrong - see below.  The corrected sim shows excessive quiescent current - either the amp's intended to run rather warm in class AB or there's something wrong with the O.P's schematic

Due to the action of D3, R13, Gyro's suggestion of inserting a 100R trimpot between R11 and R12 will *NOT* provide a useful bias adjustment range.
« Last Edit: February 13, 2019, 12:41:44 am by Ian.M »
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Trimpot rating when in place of two resistors??
« Reply #11 on: February 12, 2019, 10:14:27 pm »
@Zero999: Unfortunately 'R10' is already on the original schematic, its the emitter resistor of the current sink Q4.
Of course, the designator should be different.
 

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #12 on: February 12, 2019, 11:02:24 pm »
@Zero999: Unfortunately 'R10' is already on the original schematic, its the emitter resistor of the current sink Q4.

@All,
Please note the PMD16K100 and PMD17K100 output transistors are complimentary Darlingtons.  I didn't realise that originally, so my estimate for the max bias voltage across the Vbe multiplier Q6 was wrong.   Due to the complexity introduced by the bias limiter (D3, R13), and the varying Vbe drops with quiesent current, I didn't fancy trying to solve the bias cirucit analytically so I have knocked up a quick rough sim of it in LTspice with only one pair of output transistors and default transistor models with only HFE tweaked to get the output transistors in the right ballpark for power Darlingtons of the specified types.  Component designators correspond to the original schematic.

It shows a bias point of approx 4V across Q6 and 6.6mA quiescent current per pair of output transistors.

Due to the action of D3, R13, Gyro's suggestion of inserting a 100R trimpot between R11 and R12 will *NOT* provide a useful bias adjustment range.

Sorry gents, you have lost me a little here, my knowledge is definitely on the up but not quite there yet......

Am I to assume that @Ian.m's original suggestion of reinstating R12 and install the trimpot in place of R11 is still the correct thing to do??

And if so, was my depiction of the schematic the correct way to install it?

Thanks

Thanks.
 

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #13 on: February 12, 2019, 11:38:26 pm »
Why do you think it *needs* a trimpot and how do you plan to adjust it if you fit one? 

The addition of R13, D3 to the standard Vbe multiplier circuit adds a bias stabilisation feedback loop, so the bias should be automatic unless any part of the output stage and its bias circuit is defective or seriously out of tolerance.

The circuit currently has a trimpot in place of R11 & R12, so I was just planning to replace it like for like while overhauling the amplifier. However, when getting advice from various sources I have been told that a trimpot for bias adjustment is fine but that I should also reinstate one or the other, or both of the substituted resistors.

I have a multimeter, a function generator and an oscilioscope so I was planning on testing the amplifier thoroughly before connecting it up to my hifi system and as I understand it, the DC offset of the output can be adjust, if needed using the trimpot.

One thing to note is that while the amplifier is roughly true to the original schematic, it is not exactly the same with some components being of higher values, so I would assume that the original builder of the amp put the trimpot there for a reason?
« Last Edit: February 12, 2019, 11:45:15 pm by Robbie010 »
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: Trimpot rating when in place of two resistors??
« Reply #14 on: February 12, 2019, 11:56:38 pm »
OOPS!  |O  I missed some flyspeck decimal points in front of the emitter resistor values.  I dislike schematics that don't follow the convention of replacing the decimal point with the S.I. multiplier letter, (with R for unit ohms) and *HATE* ones that use decimal points without a leading zero.

I *SHOULD* have noticed 68R was excessive for power amp emitter balancing resistors.  0R68 is a lot more reasonable.

Please amend the values of R16,R18 before running my sim.

More to follow when I've run some parameter sweeps.
 
The following users thanked this post: schmitt trigger

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: Trimpot rating when in place of two resistors??
« Reply #15 on: February 13, 2019, 12:37:38 am »
It looks like the upper resistor should be in the 1.5K to 2.5K range.  If you've got a 10K multiturn pot, patch that in as R11 in the original schematic.   Initially set it to zero before power-on, and with no load and no signal check the quiescent current by measuring the voltage between the NPN and PNP output stage emitters.  Turn it up till you get 100mA (136mV)*, switch off and note the resistance value. Turn it back to zero and repeat to get the resistance when the quiescent current first starts to rise (1mV between the emitters). 

Once you have that data, tell us and we can suggest resistor values and circuits to allow you to trim the bias over a reasonable range while applying a low level 1KHz sinewave input and monitoring the crossover distortion with the amp's normal load connected.

* 100mA should be safe as long as the output transistors are properly heatsinked.
 

Online Gyro

  • Super Contributor
  • ***
  • Posts: 10173
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #16 on: February 13, 2019, 08:35:17 am »
Well if the OP wasn't confused before (he was)...  ::)
Best Regards, Chris
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: Trimpot rating when in place of two resistors??
« Reply #17 on: February 13, 2019, 09:52:21 am »
Yes and so was I, as can be seen above.  :-[  Without knowing what other hacks are in the original build and not on the schematic,  :-DMM usage is required to avoid  :-BROKE with loss of magic smoke.

O.T. this forum needs a magic smoke icon!
 

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #18 on: February 13, 2019, 11:55:33 am »
Yes and so was I, as can be seen above.  :-[  Without knowing what other hacks are in the original build and not on the schematic,  :-DMM usage is required to avoid  :-BROKE with loss of magic smoke.

O.T. this forum needs a magic smoke icon!

If it helps, I have drawn up the full schematic in LTSpice (to the best of my ability) and added known values and ratings where possible. Unfortunately, I am not skilled enough to run this as a spice model but it does represent the full amplifier circuit.

Attached.
 

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #19 on: February 13, 2019, 02:03:14 pm »
Sorry, just thought I may as well provide the original DIY instructions for building the amp.....
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: Trimpot rating when in place of two resistors??
« Reply #20 on: February 13, 2019, 03:01:50 pm »
OOPS!  |O  I missed some flyspeck decimal points in front of the emitter resistor values.  I dislike schematics that don't follow the convention of replacing the decimal point with the S.I. multiplier letter, (with R for unit ohms) and *HATE* ones that use decimal points without a leading zero.

:-+
Same feeling here.
Even though most schematics are nowadays CAD drawn, and therefore any distributed copies will faithfully preserve all details, there will be the time when the schematic will be printed on a low-toner printer, and similar mistakes will happen.
 
The following users thanked this post: Ian.M

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: Trimpot rating when in place of two resistors??
« Reply #21 on: February 13, 2019, 08:03:53 pm »
Those instructions say its meant to run in class A with 1V between the output transistor emitters.  That corresponds to 735mA per output complimentary pair, and a total quiescent current of 1.47A.  In practice 10% either way should be acceptable.

I'd suggest checking the current source by measuring the voltage across R10 - it should be 0.6V for 6mA current, and checking that all the resistors in the Vbe multiplier and output stage are as per original design, with *NO* potentiometer then following the published instructions from 'Final Tests' on page 6 onwards.   

If you need to mod it for class AB operation because its heating your room too much in the summer, do the tests I described in reply #15 and report back here.  You'll probably have to increase C3.  I wouldn't recommend pure class B, but you could certainly trim it down to only a few watts quiescent dissipation in the output stage.
 

Offline Robbie010Topic starter

  • Contributor
  • Posts: 35
  • Country: gb
Re: Trimpot rating when in place of two resistors??
« Reply #22 on: February 13, 2019, 09:00:28 pm »
Those instructions say its meant to run in class A with 1V between the output transistor emitters.  That corresponds to 735mA per output complimentary pair, and a total quiescent current of 1.47A.  In practice 10% either way should be acceptable.

I'd suggest checking the current source by measuring the voltage across R10 - it should be 0.6V for 6mA current, and checking that all the resistors in the Vbe multiplier and output stage are as per original design, with *NO* potentiometer then following the published instructions from 'Final Tests' on page 6 onwards.   

If you need to mod it for class AB operation because its heating your room too much in the summer, do the tests I described in reply #15 and report back here.  You'll probably have to increase C3.  I wouldn't recommend pure class B, but you could certainly trim it down to only a few watts quiescent dissipation in the output stage.

Thanks Ian.

I did read in an other forum that the design of the circuit can be altered to A/B rather than pure Class A, so I assume that is why the original builder of the amp installed the trimpot. I have no interest in that so will convert it back to standard as per your advice.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf