MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

STG MIXER (CP3 clone) - POWER SPECS? (not published)
MUFF WIGGLER Forum Index -> 5U Format Modules  
Author STG MIXER (CP3 clone) - POWER SPECS? (not published)
Rex Coil 7
Anyone know the power requirements of the STG Soundlabs "Mixer" module in 5U?

I've checked the website, and nothing is mentioned (which seems either odd or negligent). I've looked carefully at the PCB itself. Nope.

I've emailed STG, but no reply yet (been two weeks). PM-ing him is not working either, he hasn't logged here in for two weeks.

So .. anyone with any knowledge of the power draw per rail on the STG Mixer? I have one, I enjoy it, but I needs me some specs. I'm configuring a new power distribution system, so the need for these figures has come about.

Thank you! smile
goom
This lists the MU version at +/-15V, 10mA per rail:

https://www.modulargrid.net/d/stg-soundlabs-mixer
Rex Coil 7
goom wrote:
This lists the MU version at +/-15V, 10mA per rail:

https://www.modulargrid.net/d/stg-soundlabs-mixer
Thanks for responding.

Hmmm ... 10 milliamps per rail. That seems kind of "light". For instance the Dot Com Q112/Q113 mixers are listed on the Dot Com website as drawing 30ma per rail (3x more), and those two mixers don't produce any voltage offset (the STG has two channels that do double duty). I'm no engineer (not by a long shot) but it seems like 10ma per rail is incorrect. If it is correct, then good for STG for creating such an efficient module!

I've seen some ~goofy~ specs on Modular Grid sometimes. I don't use the forum or the virtual-systems-feature there (I just use it as a way to look at modules in a database that provides Function Indexed reports, so if I want to see a bunch of different ... let's say ... VCAs, I'll use it for that). But it seems like some of the specs are entered by end-users many times. Makes me wonder where they got the information in the first place.

I don't know what to make of that spec. hmmm.....

Well, I suppose I can simply presume that the STG Mixer draws (picking a number from the sky here .. overshooting to be on the safe side) ... oh ... I don't know ...... 40ma per rail. Just to err on the side of caution.

Ok, thanks for the help! I appreciate you going out of your way to locate that for me. At least it's something (which beats nothin')!

thumbs up
ranix
on this topic, is the rule of thumb in modular land not to exceed 80% of your power supply's maximum rated load? Or does the average modular power supply expect you to max it out?
Rex Coil 7
ranix wrote:
on this topic, is the rule of thumb in modular land not to exceed 80% of your power supply's maximum rated load? Or does the average modular power supply expect you to max it out?
Many (most?) power supplies can't handle a full load and still operate at 100% duty cycle. Most offered these days can barely hold full load at 60% duty cycle, let alone 80%. Not to mention that many PSU vendors pad those specs. They often times don't actually provide their advertised load on EACH RAIL all at once ... it's usually a total figure that must be shared by all three rails.

EG; PSU "X" specs out at (let's say) 1500ma ... most times that actually means 1500ma divided up among all three rails (+12v .. -12v .. +5v .. combined). Or it means that it can actually provide that kind of power to all three rails, but only for a few seconds, perhaps one minute, out of every ten minutes.

So when a PSU ad says the PSU is rated at 1500ma on all rails, they're not lying, they just haven't told you the whole truth (that's the same in my book). It will deliver 1500ma on the +12v rail, and it will deliver 1500ma on the -12v rail, and it will deliver 1500ma on the +5v rail ..... just not 1500ma on all three rails all at once ... or for full load periods without intermittent cooling periods every few minutes (this is known as their Duty Cycle).

Try to find PSUs that deliver rated loads on all rails simultaneously at 100% duty cycle (full rated power on all rails at the same time for indefinite periods .... they're not quite unicorns, but they are not usually found in the most conspicuous sellers' sites either). For instance, an honestly rated PSU that's rated at (again, let's say) 1500ma on the +12v ... while being able to deliver 1500ma on the -12v ... while being able to deliver 1500ma on the +5v rail .. all at once, for 10 minutes out of 10 minutes (no cooling period required). Some PSUs CAN deliver maximum rated loads on all rails all at once, but not for ten minutes out of ten minutes ... they can only provide those types of loads intermittently, with cooling periods of reduced loads in between the total output periods. Or offer max loads on all three rails, but not on all three at once (1500ma "divy'd up" between all three power rails).

Sellers will pad these specs with sales copy that goes something like "1500ma of raw power on all rails!" ... yuh ... sorta ... that much total power IS available on any of the rails ... just not on all of the rails, at the same time, or ... on all of the rails at once but not for 10/10 minutes ... there must be periods of reduced power output to cool the windings in between short term full load runs. Naughty naughty!

This is why this "80%" rule of thumb has become popular doctrine. It's actually more like 60% to be honest. There are few and far between PSUs that can actually operate at 100% of their rated loads on all rails all at once. There ARE some, but usually not the popular ones.

At risk of being labeled a shill, I can say that when I think of one of the honestly rated power supplies, Hinton Instruments "FullPower2" PSUs come to mind. They are rated at full load on all rails for indefinite periods (full power, all the time, on all rails at once). No BS, just straight up dead honest specs.

It's not that other PSUs are inferior, it's just that marketing departments fudge on the specs or use hokey language to be able to sell less powerful PSUs marketed as more powerful PSUs to appear competitive. The poor synth customer trusts the vendor (after all, they are the well known "So-N'-So" .. they'd never take advantage of me) ... that trusting synth customer just wants to power up his/her synth, they don't want to have to complete a degree in engineering to be able to properly spec out their own PSU. Some vendors exploit that.

Rather than running your PSU at 60% of it's rated load ... just buy a more powerful PSU that is honestly rated. It really is that simple. It really is! cool
Paradigm X
i dont know the stg version but i built a Manhattan analog cp3 clone and 10ma wouldnt be surprising - theres only a few transistors and passive parts in total.
daveholiday
I would think that any decent "capable" test meter should tell you what you need to know. I have done many amp load checks in various scenarios, but not on my modular. The only "hard" part of the test is finding the proper point to run the circuit (each rail) through the meter. Granted you would have to test each rail separately, but this should tell you real world working load. I am not an expert at this stuff either, but hopefully some of the veterans will correct me if this can't/shouldn't be done.
Just me
daveholiday wrote:
I would think that any decent "capable" test meter should tell you what you need to know. I have done many amp load checks in various scenarios, but not on my modular. The only "hard" part of the test is finding the proper point to run the circuit (each rail) through the meter. Granted you would have to test each rail separately, but this should tell you real world working load. I am not an expert at this stuff either, but hopefully some of the veterans will correct me if this can't/shouldn't be done.


The easy way in MU is to build a breakout cable. Plug the power connector into the cable, connect a flying lead from the breakout device to the module.
use the test points on the breakout device for metering.
Rex Coil 7
I am adept at current and load measuring, I owned and operated a welding machine and auxiliary generator warranty, repair, and installation shop from 1994 to 2009 (I sold the business due to an inoperable neck injury getting in the way). 18 employees, 7k sq/ft shop, three 10-kilowatt load banks, and generated 1k invoices per year. Service contracts with over 40 major manufacturers of power equipment as a Factory Authorized Warranty and Service Center. Everything from power tools, welding machines, plasma cutters, RV and portable generators, oxy-acetylene torch and regulator repair, on up to 1 Megawatt emergency generators. I still have most of my current clamps for my meters, however my favorite Fluke 860B was stolen from me by someone I was interviewing for a job last year. I haven't replaced it with anything decent, been using a couple of $10.00 meters for the time being until I get around to uprating to better gear. It simply hasn't been a priority.

But, the suggestions regarding how to measure current draw will, I'm certain, be highly useful to others that read this thread. So, BIG THANKS to those that offered the process! That was thoughtful of you who chipped in.

But seriously, I don't need to know that spec THAT badly. It's not worth it to me to go to that kind of trouble just to find this out about the STG Mixer. This should be a published spec, to be honest. I don't think I've seen a module built by a serious manufacturer that didn't have current specs available. It's something that modular synth builders require to be able to properly add up the total load of their modules to properly assess how much power their synth will require before they buy their power system. It's also something that is required a lot of times for troubleshooting problems, both on a module itself, as well as a system problem.

It's not that big of a deal, really ... I guess. If, in fact, it really is only drawing 10ma, I can just use an over-shot estimate of current draw (say ... 30ma?) and take things from there. After all, we're only talking about a few milliamps relative to a 1 or 2 amp PSU.

I just needed to know if it pulled a lot more than your buck-standard mixer module. It looks like it may not. Yay!

Odd how this spec just seems to be unknown and unavailable. Oh well, I guess I'm expecting too much.

Once again, thank you very much for the suggestions. I appreciate the sentiment. I'll just make do without, no biggy. nanners
Gizmo
Rex Coil 7 wrote:
Anyone know the power requirements of the STG Soundlabs "Mixer" module in 5U?


Your question prompted me to build a breakout rig -- suggested by a post here but on my to-do list forever. Measurement of my STG mixer is 13-14 ma on each 15V rail and zero on the 5V rail. It doesn't seem to vary whether operating or quiescent.

EDIT: The above 15V figures are wrong (see later post). Correct figures are around 27mA each 15V rail.

Rex Coil 7
Gizmo wrote:
Rex Coil 7 wrote:
Anyone know the power requirements of the STG Soundlabs "Mixer" module in 5U?


Your question prompted me to build a breakout rig -- suggested by a post here but on my to-do list forever. Measurement of my STG mixer is 13-14 ma on each 15V rail and zero on the 5V rail. It doesn't seem to vary whether operating or quiescent.

Look at you go! Right on, thanks for taking the time to obtain this. And good on ya for measuring it in both conditions. Well done.

I've made note of your findings, and I appreciate your effort.

Thanks once more. thumbs up
goom
Hey Gizmo,
In your pic, your meter probes are hooked up to the meter strangely. Did you do the current measurement with the probes hooked up like that?
goom
I'm also curious about the breakout rig connections and current measuring procedure...
Rex Coil 7
goom wrote:
Hey Gizmo,
In your pic, your meter probes are hooked up to the meter strangely. Did you do the current measurement with the probes hooked up like that?


goom wrote:
I'm also curious about the breakout rig connections and current measuring procedure...
He used those terminal blocks to connect two power cables to live input power and the module. In other words, he connected his meter in series with the module and it's power input cable. So it goes ...

POWER ---> open connector ---> meter probes (one connected to input power, the other connected to the "load", which is the module) ---> to the other open connector ---> and finally into the module. This puts the meter in series with the input power as it flows into the module. All of the power the module draws has to flow through the meter to power the module (the meter actually becomes part of the circuit). When you measure voltage, your meter is in parallel to the voltage test points, no voltage has to flow through the meter to complete a circuit.

So, once more, to be clear:

Power ---> meter ---> module.

The meter then measures the amount of current that is flowing through itself as power flows to the module. If you disconnect the meter from this setup, no current makes it to the module, you've opened the circuit.

Regarding the way the probes are connected to the meter ... the black probe is connected to a jack that is normalled to the "COM" jack, which completes the circuit in the meter. (look at the white lines that run from the "COM" jack to the "ma" and "10A" jacks, indicating those two jacks are to be used as "COM" when measuring current. You don't connect the probes to "COM" and "ma" ... those two are essentially the same connection point. Plugging in to "ma" or "10A" is .. functionally .. the same as plugging in to the "COM" jack. The difference being the "COM" jack is used for measuring voltages, and the "ma" and "10A" jacks are for measuring current. Think of those three jacks as being "multed" together to "COM" when you picture how it works.

It's correct. thumbs up
ranix
That does not appear to be correct. From the Fluke Series II manual:



I haven't used a Series II but on the ones I have used current measuring is done as pictured.

Is it really normalled on this meter? That would be really weird. If so, I wonder why they changed the layout.

Actually the one in this manual has different markings on the front than the one pictured.



Can't find the manual for the original 75
Rex Coil 7
ranix wrote:
.....Is it really normalled on this meter? That would be really weird......
Figure of speech ... I was attempting to draw a mental image for the reader. They're not really "normalled". I'll just say this ... Flukes can measure current. That's a pretty safe statement, I think I'm comfortable with that one.

To your question ... I have ZIP CLUE. Apparently that happens to me quite a bit. hmmm.....

The photo may have been taken in haste, perhaps the person didn't connect it properly just for the photo, perhaps his meter is different, perhaps he's lying about actually testing the circuit at all. I suppose with all of those potential variables, it would be best if I were to just permit the person that posted the pic to answer for himself. hmmm.....

I'll learn one day. Uh.... maybe. meh lol lol

(This is self mockery ... please take my post as such ... just poking fun at MYSELF). Dancing Star
Gizmo
Well I screwed up. Without the photograph I would not have believed it. But I just repeated the measurements with different results, so I have to conclude that my meter connection was wrong and therefore my reporting was wrong.

So the correction is: Around 27 mA per 15V rail and zero on 5V. Considerably higher than expected. Photos below.

My apologies to the community, and hope you don’t hold this against me. Thanks to the alert members who noticed.

Rex Coil 7 wrote:
The photo may have been taken in haste, perhaps the person didn't connect it properly just for the photo, perhaps his meter is different, perhaps he's lying about actually testing the circuit at all.

I would never falsify data. It is contrary to my principles and my nature. I do know how to make current measurements and this was an instance of carelessness – and I am careful in the extreme so, to me, this is as astonishing as it is embarrassing.

Something insidious at work here was confirmation bias -- I was expecting in the vicinity of 10 mA and “found” it reasonably, so there was no reason to doubt my work.


goom
No worries... we're all here to support ourselves as a modular synth community. smile

I'm glad we were able to find the answer to Rex7's original post. I was curious too. smile

I don't have an STG CP3, but I did build my own based on the original Moog schematic. I'm sure mine draws less current than the STG version, as I don't have any voltage regulators, or other additional components in the circuit (compared to the original).

Rex Coil 7
Gizmo wrote:
Rex Coil 7 wrote:
The photo may have been taken in haste, perhaps the person didn't connect it properly just for the photo, perhaps his meter is different, perhaps he's lying about actually testing the circuit at all.

I would never falsify data. It is contrary to my principles and my nature.
Please, don't take that personally. I was pragmatically pointing out that there are always any number of variables in a given situation that involves human beings. In my experience in dealing with situations involving us humans, I automatically build-in a ~roughly~ 20% handicap for failure .... it's not because I do not have any faith in people, or that I'm cynical, it's a matter of acknowledging that humans are not perfect. There has never been a flawless human, and there will never be a flawless human (a flawless human would not be .. well .. human ... a flawless being would be something other than a human). That said, it's simply wise to give people a "pass" when it comes to failure or other-than-optimal performance.

As far as considering the possibility of deceitful intent, it wouldn't be the first time someone had made a false claim on (insert dramatic reverb here) THE INTERNET. eek! eek! DUN Dun duhhhh.

I don't "think" that you are dishonest, nor do I "feel" that you are dishonest, my comment about the very real possibility of deceit (just as likely as any other failure point when dealing with someone ... until you get to know them and they demonstrate their genuine efforts) ... not because ~you~ .... specifically you, Gizmo ... are JUST a dishonest person ... that word "just" highlighting that dishonesty is your most noteworthy personality trait.

Keep in mind what I said at the end of that very post ..... "this is self mockery ... I'm making fun of myself". I was essentially throwing my arms in the air, acknowledging the fact that Goom had (quite correctly) pointed out that I WAS WRONG in my support for the manner which the meter was used. I had made a fool of myself in making some professorial posting about how to ~ahem~ "properly" connect a meter to measure current and in my almighty wisdom had assessed that you (Gizmo) had done it properly. I was aiming a truckload of sarcasm AT MY SELF, pointing out that I was obviously the wrong person to answer Goom's inquiry ... and making a point that Goom made the proper pragmatic assessment of the image of the test connection, and I had failed to take everything into consideration.

In other words .... in frustration AT MY SELF ... I was spouting the sentiment of "don't ask ME, I've obviously missed the obvious and should not be relied upon to provide an accurate confirmation when my brain was ... at that time ... failing the hell out of not only myself, but anyone within the Muffwigglers forum that may have come to trust my judgement. I'm messing this up so badly right now that I may have missed ANY NUMBER of things ... har har har I'm so stupid har har har ... if anyone would like to sell me a bridge, NOW is the moment since I clearly have my head up my behind today" ..... THAT'S where that "lying" comment was coming from.

CLEARLY, a hoof-in-mouth disease moment for myself. I mean, you might have made a very human error ... but I clearly STEPPED ON MY DICK! (That means I was over-confidently complacent, and because of that I failed to see how wrong I was, the result of which caused pain in the form of humiliation).

So please do not take my comment personally. It was just one of those days for me that the only time I open my mouth is to change feet.

ON A DIFFERENT NOTE:
AH-HA! So my instincts were correct about that mixer .... "10ma" (or thereabouts) just didn't sit well with me. And the fact that "someone" (who?) posted the 10ma figure on Modular Grid was not to be taken seriously until that spec was either confirmed or impeached by testing, or through verification by the circuit designer and/or builder them-own-damn-self.

I'm just SUPER GLAD that you .... Gizmo ... in fact ARE a person of such integrity that you're willing to take responsibility for yourself and post the correction. Because now the correct current draw of the STG Mixer is known ... and published! (Y'know, that said, this whole thing could have been avoided if the specs were published in the first place ... both of us would have less egg to wash from our faces).

I'm not tellin' I'm just sayin'.

27ma +/- it is then (I'm still going with 30ma as a rounded up figure, from a safely cautious position). I don't feel as though that observed amount was "higher than expected" at all, given that many other mixers draw roughly that as well.

Very well ... and thank you for your efforts, Gizmo. And to be clear .... I APOLOGIZE for offending your sensibilities, Gizmo. My comment was inconsiderate.

I'm classifying this entire affair as a learning moment.

Brian.

cool
Gizmo
Rex Coil 7 wrote:
Please, don't take that personally...

I did not, and no apology is necessary. Actually I thought you were very generous in listing so many ways of giving me the benefit of the doubt. My reason for responding on that one point is that our online reputation hangs crucially on credibility, since we have no in-person or other clues to rely upon and judge by.

And you are right about lessons learned. The damnedest one for me in all of this is that I made this blunder despite several decades' worth of electronics tinkering and measurements. Plus I am obsessive about mistake-proofing. I carefully laid-out the power connections in a consistent order and used color-coded jumpers precisely to prevent power cross-ups, because that (in the words of Egon Spengler) "would be bad."

So matter addressed and now back to using all this stuff.
Gizmo
goom wrote:
I don't have an STG CP3, but I did build my own based on the original Moog schematic.]

Nice. Compact, complete and good-looking. Do you get the expected audio character?
Graham Hinton
Measuring the current by inserting a DMM in the path is not accurate because of the meter's "burden voltage" which is a voltage drop across the meter's internal resistance. This will be buried in the smallprint of the spec and some meters are a lot worse than others. It may be enough of a drop to make the module function differently.

http://www.fluke.com/fluke/uses/comunidad/fluke-news-plus/articlecateg ories/electrical/burdenvoltage

A better method is to insert a very small known resistance in the path and measure the voltage drop across it and then use Ohm's Law, I = V/R. Very small being something like 0.1 or 0.01 ohms, but even than is too high to leave permanently in place as it can cause module coupling.

The best method is to measure the magnetic field around a wire carrying the current. We have been using this technique in our Power Monitor option for a couple of years and from this autumn it will be standard in all our PSUs. Although this is used to show PSU overload on LEDs, the raw detector voltage may be viewed on an oscilloscope to show the actual dynamic DC + AC current.
goom
Gizmo wrote:
goom wrote:
I don't have an STG CP3, but I did build my own based on the original Moog schematic.]

Nice. Compact, complete and good-looking. Do you get the expected audio character?


Thanks for the compliment!

I do like the sound of it, and use it often. I do wonder how close it is to the original module in terms of output. One of these days, I'll hook a scope up to see what the output actually looks like at various gain settings.
Gizmo
Graham Hinton wrote:
Measuring the current by inserting a DMM in the path is not accurate...

Right. This is not news. Any invasive measurement changes the system under measurement in some way, besides being subject to inherent inaccuracy. The question is, is the measurement system accuracy *good enough* for the intended purpose? For the case at hand, within the broader goal of managing system power, do we have accuracy good enough for the purpose of knowing a module’s consumption reasonably?

The burden voltage for the Fluke 75 is specified as 6mV/mA. Assuming 15V supply rails and an actual 27 mA current, the burden voltage drop is 0.162 V, or a little more than 1%. Is this likely to affect the STG Mixer behavior in a material way? I am not going to assert that it does not, but I have the opinion that it is not a significant effect.

There is the matter of the rated accuracy of the meter itself. The Fluke 75 when new has a published accuracy in the low current range of ±(1.5%+2) which, assuming an actual current of 27 mA and a two-decimal-place display, works out to ±0.425 mA, leading to readings from 26.575 to 27.425 mA. Still, that performance level applied only when the meter was factory-new; though it has served me well over the years, it has likely worsened with time.

There is no question that differences in the operating parameters -- in this case, currents of several hundred milliamps, for example -- would strain the accuracy and precision assumptions and suggest more careful investigation and greater caution.

In formal assessments of measurement systems there are many arcane impairments than can come into play, reaching for example to the ambient conditions of measurement, warm-up times, operator competence (“reproducibility”), instrument variations (“repeatability”), measurement system drift over time, and on and on. Even, in our case, whether we can say that my STG Mixer module is typical of the general build population. Maybe my unit’s components are unluckily skewed in one direction or another, or some other factors affect its functioning, so that my unit is not representative

No doubt it is prudent to exercise caution when making and using measurements, which is constructive and valuable advice. Those of us who have used electro-mechanical instruments -- ancient dials with moving needles -- learned early to be skeptical of the readings and take steps to compensate or otherwise allow for errors. (Anyone here remember mirrors behind the needle to reduce parallax error?)

So here I am going with 10% and am comfortable (let’s say for giggles to the .05 level of confidence, though I have not done a Gage R&R) that the actual current value is within that range. Readers here are free to ignore my measurements and/or perform their own to their satisfaction. I have shown what I did and how I did it; anyone can thereafter judge whether it fits their intended purpose. For me, this exercise fills-in a gap in my own power consumption worksheet at 30mA, which is quite good enough for my purposes to manage my power supply and distribution.
MUFF WIGGLER Forum Index -> 5U Format Modules  
Page 1 of 1
Powered by phpBB © phpBB Group