Bench power supply recommendations / talk me out of AIM TTI please

[solar_pi]

Hi all, I am a lurker till now and a new EEV community registered member.

I have been gradually doing more tinkering, soldering, Arduino, LED and RC stuff  (with cheap PSU or LiPos).  I finally decided to pull the trigger on a decent Bench PSU. Over the last 2+ weeks I spend more time than I am willing to admit doing my research and was about to pull the trigger on PL303QMD-P, but put it on hold (see below).

This purchase will have to last me for years if not decade to come.

I am looking for:
- at least 2 channels (i.e. 1 for Arduino, second for LEDs / in the future - parallel +/- voltage)
- 0-30V on at least 1 channel
- 3A at least on 30V channel / 90W power on said channel
- decent resolution and accuracy
- ideally - old display style / no TFT or OLED screen

Here comes points I am not sure how to value
- remote sense - how important is it given I might be using longer leads in the future.
- activation / reaction speed of OCP/OVP - how fast does it have to be?
- programmability (not yet needed, expect to need in the future <<i.e. for LabView which I used as a student>>, when I don't want to buy another PSU)

My original favorite was AIM TTI PL303QMD-P until I read that reaction time on OCP/OVP is 500ms (!). Is it just me or does it look like it is very long (especially by comparison to the likes of Keysight products which claim to have 1.5ms or Rohde and Schwarz who claim 10ms).

Bench power I considered:
- BK Precision 9129B/9130B, Keithley 2231A - seems to be same product. Some info on this forum and on YT seems to be showing less-then-optimal design for the price
- Rigol DP932 - actually very bulky (I have limited space on my desk), no remote sense, limited info on OCP/OVP. TFT screen
- GW Instek GPP-3323 - no details on OCP/OVP. TFT screen
- Siglent SPD3303X - I believe Dave already took it off my list
- AIM TTI PL303QMD-P - still potentially on the list. QL II version is way beyond my budget, so would prefer to stick to PL series.

Given the above and your own experience with said or other Bench PSU I would greatly appreciate if you took me out of my misery in search for the right PSU so I could focus more on my little projects again!

[ledtester]

Don't spend USD $800 on a power supply. There's so many other things to spend money on in this hobby, e.g. a second (or third) multimeter, function generator, calibration reference, thermal camera, microscope, Analog Discovery, electronic load, battery powered scope, high voltage probes, nanoVNA, tools, parts, pcbs, ...

Over the years I've accumulated a number of linear power supplies all of which I bought used and paid an average of around $50 each for. The one I keep coming back to is a Tenma 72-8335A -- a very basic, dual 24V/1A linear supply which you can tell from these videos is one that was built down to a cost:

https://youtu.be/6vEqTPi1Uzc
https://youtu.be/cnjv8ZjCtgk

CC is nice to have, but I've never used remote sensing, OCP or OVP.

One advantage of a simple linear supply is there's a good chance you can fix it yourself if you happen to blow it up.

As an alternative to a basic linear supply I would suggest looking at the switching supplies made by GopherT, e.g.:

New Gopher PPS-1610 PPS-3205 Review & Teardown | Voltlog #391
https://youtu.be/1qCrWQJldlw

[jonpaul]

Bonjour, nice to hear from Switzerland:

I have no knowledge of the Chine digital controlled supplies you mention, but as an EE since 1968 specializing in analog and power design, all of my lab supplies are old classis linear USA made  with analog meters. A dual or triple supply is essential to run 3.3, 5V digital and +/- 3.3...5...9..12..15V analog. We have several HP 6236B very high quality and troubvle free.

For higher power many HP, Lambda, Power Designs are in the V/I the OP asks.
In general analog meters are better to follow trends and adjust V or I.
All lab supplies are CV/CC and limit instantaneous current.

Finally a crowbar at the output can protect the DUT.

You will find such classic PSU at Ham Radio flea markets and ebay but beware of shipping cost.

Bon Chance,

Jon

[voltsandjolts]

We have a couple of these PL303QMD-P. They are very good but not as sturdy as the old PL series, and there is often quite a 'thunk' when you switch them on.
The user interface is excellent IMHO, with neat trick of setting a custom range for the adjustment knobs.
Fan is variable speed and very quiet at low current draw.

My original favorite was AIM TTI PL303QMD-P until I read that reaction time on OCP/OVP is 500ms (!)

OCP/OVP reaction time is the time to switch off the output, 500ms is fine and that is much slower than the current limit reaction time - I think you perhaps confused the two.

As a beginner you can get a perfectly good linear psu on ebay for a couple of hundred pounds, no need to spend more.
They hold value well so you can upgrade in the future if you really need to.

I would recommend something like a tti pl320qmd as an excellent starter linear bench psu e.g. https://www.ebay.co.uk/itm/302374067634

[metebalci]

- at least 2 channels (i.e. 1 for Arduino, second for LEDs / in the future - parallel +/- voltage)

Maybe depends on what you are doing, but if it will be your only PS, I recommend 3 channels if your budget allows / the feature list of the device fits. Multiple times I found myself in a situation thanking I have 3 channels.

[solar_pi]

First of all - thank you all very much, this is a great community. Secondly - I have seen some 2+1 (2x 0-30 and 1x 0-5V) power supplies, but they seem very expensive. 2 Few more points I have to consider:

- my desk space is very limited
- customs of getting anything into Switzerland are significant (VAT + Customs are put on top of purchase price + shipping costs, which for PSU being north of 8kg can be pretty hefty)

3 outputs seemed to be covered with BK Precision 9130B, which is a rebranded version of something else and does not get good press

also - this is what I meant by PL reaction time:
" alt="" class="bbc_img" />

[solar_pi]

Long story short - I stick with single output PL303.

I contacted TTI support today and the gentleman on TTI side was kind enough to share:
- 500ms is a reaction time to settings change for OCP/OVP in the -P version of the PSU
- QL II (which I started to consider) is a slightly older construction
- no new linear PSU from TTI is expected in the line of QL/PL

double/triple outputs and programmability, as much as very tempting, would eat my budget which I can now allocate slightly differently in the hobby!

[trp806mo]

I think you have some other points to take into account as we don't know your application :

• Ripple noise (do you need a linear one for some audio purpose or analog stuff)
  Type of outputs (1,2,3 , or 2 with always a fixed one to 3.3/5/...) ...
  Range for the V, I
  Noise of the power supply
  Output on the front/rear
  Isolation needed between the channels (some have 3 complety floating outputs some none)
  Synchronisation between chanel (independant, serial, //, +/- interesting for the AOP, tracking, synchronisation between V1/V2/V3 if used on a same board ...)
  Compacity versus price (you decrease the volume but not the price with a 3x output regarding to 3x1 output)
  weight (yep a 3xlinear has not the same weight than a dc/dc) if you need to move it often

Here my feedback (which depend of my applications) :
I used to manipulate a lot of low power power supply (<1Kw) at work or at home tti (QLxx / PLxx/EL) old HP6237B or HP66xx/Korad/Keysight/ R&S/ ...) :
I'm using only one very low ripple noise power supply for some analog stuff (but i could have used a lipo or lead battery too).
I rarely used a remote sens (only of some FPGA rails under 1.2v  with high amps). I guess you will not have more than 5/6A  and loose only some hundred of mV  in your application
I have more concerns about the usability (be capable of reading the current of all the output at the same time [LCD display mandatory] / having the possibility of configuring the output as i want (V1 after V2 after V3 or V1=-V2) handling SCPI API
The TTi (check topics) are not so well build for the price (output capacitors are no high grades on the QL/PEL series for example)
I like to use a  RD6012 with some motors as i don't care about the ripple noise and i like its lightness when i have to move it
on the day to day in 90% of my work a Korad at 50€ will do the same job as my R&D at 2K€ as my DUT have already some internal power supplies
(As advised) I prefer to put 1K€ in some other stuff than in a power supply as a Korad at 50€ will do the job
(As advised) you can buy on ebay some second hands  power supplies for 1/4 1/5 the price of a new one. For example i bought in France a QL355P for less than 200€ in a mind condition on Leboncoin (i don't like it for the noise by the way)





 

[Jeroen3]

If you only want to buy one power supply. Perhaps the GW Instek GPP-4323 is of interest.
4 channels is definetely enough, built-in series-parallel switching, 1 of the channels can be load as well. Very feature rich.
Doesn't weigh that much and isn't loud.

But it doesn't have knobs....

[trp806mo]

You have one on the upper right ... you can configure what you want with it.
the knob is mandatory when you want to vary quickly the PS output and check the behavior off the DUT, something you are not doing every day ... The keys are more convinient to input the right configuration quickly.

[solar_pi]

Ripple noise - audio
Type of outputs - so far I cannot imagine a project where I need > 2 outputs
Range for the V, I - max 30V, at least 3A
Noise of the power supply - the lower the better
Output on the front/rear - at least front
Isolation needed between the channels (some have 3 complety floating outputs some none) - not sure
Synchronisation between chanels - needed if +/- voltage would be required on outputs
Compacity versus price - compact for now is quite important
weight (yep a 3xlinear has not the same weight than a dc/dc) if you need to move it often - weight does not matter, but the footprint yes.

I would still like to understand how important remote sense is.

I did order PL303, but I can send it back in first 14 days, so if GW Instek is the next best thing after sliced breat - I could potentially use this option

[Electro Fan]

Congrats on the PL303.  It might be just what you want, but if not…. another Plan B, or C.

Buy a KORAD KA3005D from Amazon for $110.  Pretty small footprint.  Doubt that you will need remote sense.  You will probably like the Korad a lot but if you don’t for any reason you can return it.  If you like it buy another.  You will then have two channels (each up to 30 Volts, 5 Amps) with pretty low ripple.  $220 total - will save you some funding for TEA.

[trp806mo]

The PL is very compact and noiseless until you are not at 80% of its maximal power. Take only care that the airflow is going from the bot to the top and not to the rear.

For the sense you will be able to play with it It's useful when you want to compensate the loss of the cable under high current. It depends of you application. For mine, if you have a device at for example 1.2v under 5A, you may loss easily  some hundred of mV which may reset your device. It can be usefull too when you have to perform some "precise" measurement and don't want/can connect a DMM for reading the input power of your DUT.

 

[wizard69]

This is the best answer and it is right at the top.   I have to agree there are too many other things you can buy that will be more useful in a lab bring up.   As for power supplies they are all over the place if you go looking and can often be had used cheap or even as salvage.    On top of all of that building and repairing power supplies are great skill builders.

Now at some point you will want to buy a power supply but I'd try to focus on the more useful things first.  When you do get that power supply a few years of getting buy will help you really focus in on the features you will really want.    For example a port of some sort for remote programming in most cases is not a big thing in a home lab.    However if you are trying to bring a product to market it (remote programming) might be required.

If you want to know what I consider to be more important than a bench power supply, I'd make sure I had these first (no order):   

• Multimeters; at least one handheld and one bench with more in the distant future.
• A scope.
• Soldering station.
• Hand tools.
• Small drill press.
• Lab computer.
• Lighting.
• A decent bench.
• Tool box, storage solution.

.

Don't spend USD $800 on a power supply. There's so many other things to spend money on in this hobby, e.g. a second (or third) multimeter, function generator, calibration reference, thermal camera, microscope, Analog Discovery, electronic load, battery powered scope, high voltage probes, nanoVNA, tools, parts, pcbs, ...

Over the years I've accumulated a number of linear power supplies all of which I bought used and paid an average of around $50 each for. The one I keep coming back to is a Tenma 72-8335A -- a very basic, dual 24V/1A linear supply which you can tell from these videos is one that was built down to a cost:

https://youtu.be/6vEqTPi1Uzc
https://youtu.be/cnjv8ZjCtgk

CC is nice to have, but I've never used remote sensing, OCP or OVP.

One advantage of a simple linear supply is there's a good chance you can fix it yourself if you happen to blow it up.

As an alternative to a basic linear supply I would suggest looking at the switching supplies made by GopherT, e.g.:

New Gopher PPS-1610 PPS-3205 Review & Teardown | Voltlog #391
https://youtu.be/1qCrWQJldlw

[radiolistener]

Two KORAD KA3005D will cover all your needs.

[wizard69]

I would still like to understand how important remote sense is.

For the most part on a bench supply it isn't important at all.   Now if you are going to build up some fancy instrumentation, for a commercial installation some place, it could be very important.    I just fall back to the idea that a power supply isn't a high priority instrument to have on a bench that is being newly built.   In most cases almost every other item for that bench is more important.

[artag]

The intention of remote sensing is to compensate for the voltage drop along the power leads. Whether this is important depends on your application, but since a bench supply is there for *all* applications I would say it's worth having, especially on a supply that's more than 2-3 A.

I have also found it useful for another case :  you want to measure (and perhaps show on an oscilloscope) the current taken by your test device. Obviously you do this with a resistor in the lead : but if it's a small resistor the voltage drop is hard to measure, and if it's a large resistor you lose a lot of voltage and this may be important - for example when looking at the performance of a boost power supply when running off a simulated battery.

But if you put your sense resistor between output post and device, with the feedback connection connected to the device rather than the output post, the power supply will compensate for the drop across the sense resistor. You will get the required voltage at the device regardless of load, and you can have a decent signal to measure current with.

Note that power supplies will not generally tolerate a very large correction - they're supposed to compensate for lead drop, not several volts. So make sure it's working as expected and actually producing the required output without oscillation. But it's a useful technique. For a reasonally well-equipped bench, I would always want at least some of the power supplies to allow it (and all of mine do except for a low-current triple rail HP E3630A, though some require me to use the rear terminals to do it).
 

[solar_pi]

For the sake of closure - Aim TTi PL303. 30V, 3A. I feel the lack of current on occasions (4 8x8 RGB LED panels from Aliexpress can confirm), but it is such a small form factor PSU that I cannot complain!

Remote sense - did not matter a single moment so far
Small form factor - perfect if you are not rich in terms of desk space
Voltage limiting - great when working with small electronics
Lack of possibility to key in the voltage - would probably be more important if more often would I change the voltage above 5.

Would buy again: 10/10