M670 compressor

[abechap024]

Looks awesome. I would love to build one of these for the studio. Great work Kingston! Thanks for sharing. Thats awesome you've done a lot of prototyping, I know sometimes its a hard process to get rid of all the kinks, despite the best efforts. Also if your doing pcb design, might be work checking out diptrace i like it at least, though peoples mileage may vary (I could never get into eagle) count me in if you do a PCB!

 

[Kingston]

As far as I remember there was a quite long discussion before the pcbs were released. Analag had a prototype.

I'm well aware of this. It's the usual thread with people salivating over an imminent PCB release. A little bit light on design data! Besides a trivial PSU transformer specification, here are the only design related comments, which are actually rather important, considering this new M670 rework.




In chronological order:

I'm sure something other than XSM can be used for input, I chose XSM for the output because of a number of reasons. DCR which is lower because of bigger wire, which translates to lower voltage drop and higher current carrying capacity and to a lesser degree, less effect from DCR imbalance, core size which is always a plus when DC flows through a transformer even though the pushpull action nullify the effect, higher inductance and it costs only $2.00 extra.

The unbalance in P-CT-P across the primary, is not a problem. What we need is inductance balance from P-CT-P which we have.

I'm curious to see how all these measurements compare to my own.

analag

Here's first hint that the B+ center tap at primary was a design choice that once worked. Perhaps in some other design. It was a well meaning assumption for the to-be "slimmed down, poorman version". At this point untested.

Unlike the Altec 4xxx type mu compressor the Fairchild topology is much faster, so in shortest attack/release you will hear it at the onset of compression. This can be dialed out by simply giving it a slower attack. I could have designed it out but I decided to give the user the option of the fastest attack possible.
Also notice no thumping, at all. Even with unmatched tubes.

analag

That negative cathode bias moves grid-voltage/plate-current curve to a more linear section of the curve. The single juiciest bit of the design. No major tube mismatch problems in the thump area indeed. lolo-m I'm surprised you never noticed this. Cross-over distortion is not the same as thump!

But now the PCB was finally out:

Since everybody is making a big deal about " the distortion on the left channel". In design and troubleshooting...if the distortion was on both channels, then I would look for a flaw in the design...but if one channel was "OK" and the other was distorting...I would look for a flaw in that (distorting) channel.
Remember it is two identical channels, being driven at the same levels. I specify rotory switches with matched 1% metal film resistors for maximum symmetry between both halves of the differential pair. This is even more important than DCR balance in the output transformers where some people might think the focus should be.

analag

analag was well aware of the possibility of cross-over distortion, but I suppose didn't cater for it because of the "poorman, cheap" design-mentality. I personally think any design whether cheap or quick should cater for error scenarios and should give ample room for parts selection and tolerance.

Now tube matching becomes essential and somehow nobody noticed, not even analag. This is the cross-over distortion generator bit. At this point the PCB goes on market and is sold for years to come. Whooops!

I guess you ain't read the first post I made when I started this thread. This mu compressor is not sensitive to tube mis-match. I did my homework in the cathode circuit to compensate for that.

analag

Again, thump is gone indeed, compression performance will even out with mismatched tubes. Good one. But cross-over distortion stays due to the very sensitive center tapped B+ configuration. Not even two triode-halves can compensate for that. Four halves would have a better chance (statistically) - which is a common choice out there - but that's not going to help here.






There's probably like three of you out there left on this forum who might have a clue of what I'm talking about. The above bit of paraphrased archeological digging should be rather enlightening. I suppose everyone had their best intentions (ha!), but then a half-arsed PCB was released to the market and design kinks were never ironed out. Even when they were already known. The designer disappeared with no further comments. At this point it's easier to point fingers. And perhaps it was a good lesson to some.

Still, there is certain simplicity and elegance in the design. If only it worked! And this is where I enter.


This thread, it should be a design discussion, and what I get instead is a bunch of guys in disbelief and whining how their beloved compressor turned out to be a difficult stomp box. For example Gus asked a very relevant question (which I didn't see initially, sorry). I might be confrontational, but that's exactly the kind of thing I appreciate.

 

[gato]

It's obvious you've done a lot of hard work. Congratulations.
Granted there is quite a bit of "paint-by-numbers" going on on the forum, a lot of members gradually learn and take on more challenging projects. You also do some painting, and on numerous occasions have referred to it as "knitting." I like to see the various products offered through the forum as ways we can help each other with different projects. After all, its all fun.
Of course attempts new designs, and the sharing of design ideas are the most exciting aspect of the forum for me. Not everyone knows enough to contribute, but I would say that people here are extremely generous with what they do know. I myself never post because I know far too little to contribute anything valuable, but I intend to learn as much as possible and contribute as soon as I can. I do, however, see tons of members who contribute a lot of information, including design information, usually for free. A lot of people, yourself included, have taken on crazy and inspiring projects. If it were really that bad, you wouldn't bother sharing this at all would you?
I understand your complaints, but its far more productive to do things like this, and post new designs and share information than it is to criticize each other. Of course the PM comp schematic was secret forever, and I remember how it felt very contrary to the spirit of things, but back then everyone was worried about major companies lurking on the forum and stealing ideas.
Again, good work. Cheers.

 

[zayance]

Since this is his project thread i guess talking about what's wrong or right to say will just block this thread, and we should keep it clean.
All as been said i guess, even if i believe there was a nicer way to putting this out, humanwise, but it's freewill and choices and Kingston's character so...

 

[Kingston]

Since this is his project thread i guess talking about what's wrong or right to say will just block this thread, and we should keep it clean.
All as been said i guess, even if i believe there was a nicer way to putting this out, humanwise, but it's freewill and choices and Kingston's character so...

I really don't care which way or what tone. People could feast on the assault of information, ask relevant questions on the design choices now it's actually possible. Things have stayed reasonably condensed this far, only a few "where's the PCB" type irrelevancies. I'm really only interested in the engineering aspects and perhaps project history. All of my critique has been a long time coming and well deserved. The defensiveness of some folks isn't exactly surprising. People have burned a lot of money on the old thing.

 

[lolo-m]

And poeple will burn some more on the new thing...

 

[gato]

The following are questions out of ignorance to help get back on the design track and aren't critical questions:

The feedforward/feedback switching is a new design element?

Is level only controlled by the T-pad on the input?

What does PL mean? Sorry I haven't encountered this before. These are some kind of connector?

 

[telefunk]

Perhaps that's a positive sign, but what I don't understand is the disbelief at the data presented. "It's all about the sound", they say. "I want more color". Why is it a surprise you have been listening to lots and lots of distortion?

The data presented: you build one PM670 and then you did some mods. Your PM670 happened to have some weird DC offset/distortion, so you added a cap. You assumed that everybody had the same problem, although you admit that you don't know why that DC offset appears there.

So, like I said, my PM670 was checked with a scope and it is a fact that there wasnt any of that distortion you described. We put in different freq sine waves that were automated to jump back and forth from very low levels to extreme (and everything in between) constantly, no distortions, the sine wave came out as perfect as it entered the PM670.  Adding a cap would be piece of cake, but why would i do that?

I really don't care which way or what tone.... I'm really only interested in the engineering aspects

Maybe you should be bit interested of the way and the tone, as it will just take attention away from the engineering aspects of this.

So, should we try to keep this a bit more civilized?

 

[Kingston]

The feedforward/feedback switching is a new design element?

Is level only controlled by the T-pad on the input?

What does PL mean? Sorry I haven't encountered this before. These are some kind of connector?

Input level is exactly like 1176 or the big Fairchild. It's the only one you need. PL11 just means it's connected to the other PL11 on the switch. I tried to keep the schematic layout clean this way. I don't know if anyone did this kind of feedforward/feedback switch or if it's new, just that it works. I haven't seen it elsewhere. You could also take the feed-forward input from before the t-pad for some effect-like smashing, but compression sound becomes very strange.

Your PM670 happened to have some weird DC offset/distortion, so you added a cap. You assumed that everybody had the same problem, although you admit that you don't know why that DC offset appears there.

I'll let that one slide since you clearly don't understand the completely new PSU, the plate resistor configuration or what center tap even means. What's a plate resistor anyway?

Or let's see some stomp box data, THD, IMD figures and such for some weigh in your words. I don't know why the momentary cross-over distortion happens, but it's verifiable between two equally built channels whether separately used or linked together, and two very different PSU's. Two very different transformers, any tube. This casts some doubt on your oscilloscope measurements. Did you actually do them correctly, perhaps the scope has slow screen update and/or you have wood ears? This would be a good time to present some data and wood ears is common condition in these parts and nothing to be ashamed of.

But I know why your stomp box has bad overall cross-over distortion performance even with matched tubes.

Yes, you spent a lot of time and money on it. Now someone shows you it's not the culmination of perfection you thought it was. Deal with it.

You can choose to ignore this information for the second time.

 

[telefunk]

Wood ears is common condition in these parts and nothing to be ashamed of.

I can see all the changes/designs you have made, i was just commenting  on one little detail of you mods, the "moving headroom" issue.

There can definitely be issues with my measuring and my technical abilities, for sure, but my trained ears are something you don't have to doubt. As a composer I work constantly with big orchestras etc, so listening to a little compressor is...nothing.

Anyway, i'm definitely not trying to underestimate your work, just trying to have a decent conversation of this issue which I feel is curious. But you seem to take it very personal so...

I can ask my tech friend to come and make better measurements and then collect some data to compare. 

ps. my PM670 is not nearly stock, my tech friend made several mods to it, i will call him tomorrow and ask about them, myself got bored with it so i let him finish it. He's one of the best tube techs around, so who knows what he's done with it.

 

[bruno2000]

What is the software you are using for testing?
Thanks!
Best,
Bruno2000

 

[imo]

Kingston,
I appreciate the insight. I have had my PM cobbled together and have been reticent to "finish" it for many of the reasons listed.
Definitely interested in technical details and how to maximise this unit.
Ian

 

[alexc]

Now without wishing to be too argumentative or anything, now I've had some time to look at the M670 circuit a little,

With the obvious exception of some new features such as the 'feed' options, and also the reworked power supply,

the main difference here is changing the specified model and method of connection for the signal input and output transformers and also connection of the primary of control output traffo.

The modified time constant here is relatively what most have used, and some of the operating voltages a little shifted. The GR tube sub had already been discussed by Kingston originally I think, along with things like the stereo link and vu meter.

So - while I have no doubt that these things can improve performance on it's predecessor, it remains very close to it!

Along with the new features and the simplification of the psu design and build, it could well represent a real step forward in buildability and stability. And an incremental improvement in performance, tube and other matching issues allowing.

I would surely see it as a MkII version of the mighty and controversial PM660 (actual might may vary)

It would be interesting see some implementations of it one day!

Cheers

 

[alexc]

I seem to remember reading a lot about the tradeoffs and features of using vari-gm stages of transformer-with-centre-tap-applied-B+ to feed the plates directly, as opposed to via a balancing resistive network.

They apparently gave different result and much discussion was had. Have to go looking for it.

 

[vari-mu]

I seem to remember reading a lot about the tradeoffs and features of using vari-gm stages of transformer-with-centre-tap-applied-B+ to feed the plates directly, as opposed to via a balancing resistive network.

They apparently gave different result and much discussion was had. Have to go looking for it.

  Maybe it here:
http://www.groupdiy.com/index.php?topic=24961.0

 

[creal]

Hi
This new vari-mu looks very interesting.
Thank you for this new project mister Kingston, count me in for a stereo set.
I started the PM670 since two years now and it's still not run properly today.
The PM670 project have no support, no informations and schematics for troubleshooting. it's sad...
So, after reading these posts, i consider to definitively stop to work on it and wait your new release.
Congratulation and thank for your work

Cyril

 

[bernbrue]

The PM670 project have no support, no informations and schematics for troubleshooting. it's sad...
Cyril

Please stick to the truth, there is a very detailled support thread for the PM670 with all important information to build the unit. You just have to read it.

http://www.groupdiy.com/index.php?topic=28274.0

We put very much efforts and time to share all the information we had for other users. So don´t slap people into their face when it´s simply not true. Lets go on with discussing Kingstons new compressor.
regards
Bernd

 

[creal]

The PM670 project have no support, no informations and schematics for troubleshooting. it's sad...
Cyril

Please stick to the truth, there is a very detailled support thread for the PM670 with all important information to build the unit. You just have to read it.

http://www.groupdiy.com/index.php?topic=28274.0

We put very much efforts and time to share all the information we had for other users. So don´t slap people into their face when it´s simply not true. Lets go on with discussing Kingstons new compressor.
regards
Bernd

Bernd,
Yes, i read all posts of this thread, and i like to thank you and all the other members for sharing their experiences.
It was very useful , but my pm670 still in progress and not useable in the studio.
I have buy some NOS and expensive matched tubes for this build.
Today, the compression work well but i still have a ugly noise at the outputs.
I just regret, Analag don't share the schematics and don't give any feedbacks and supports needed for troubleshooting by himself.
Is it a choice or it's too busy? i don't know ...

Fortunately, it's not the case for others gurus and master of works in this nice group.
I'm member and 'addict' of groupdiy since four years now. i finished a lot of great projects for my studio and others with sucess and very good result.
But, I still learning at every new project with help of sharing people their knowledges.
It's just my humble opinion of newbie...

 

[Kingston]

So - while I have no doubt that these things can improve performance on it's predecessor, it remains very close to it!

Along with the new features and the simplification of the psu design and build, it could well represent a real step forward in buildability and stability. And an incremental improvement in performance, tube and other matching issues allowing.

I would surely see it as a MkII version of the mighty and controversial PM660 (actual might may vary)

Fine, I'll keep this new proto to myself then, since it will always be just "MKII". A variable load DOA sidechain with threshold, which further simplifies the PSU and improves performance and safety, needs one less transformer, has an easy to use tube matching rig incorporated to it. But like you said "remains very close to the original". Since apparently all this design is as simple as "adding a cap" looks the folks here can take over. As a bonus I no longer have to shift through datasheets, calculate bias points and loads or measure a thing.

If only somebody had told me earlier it was this easy!

 

[emrr]

I have to say I was not impressed with the level of information or support on the original project, especially given how popular it was, and how many were sold. Many pieces of core information only came out after persistent prying and questioning.  I can't personally imagine offering a project and letting so many big questions hang. Anyone interested can read the original thread for evidence.  My many alternate build observations were criticized as being off-topic, thus the separate pimp/mod thread.  I'll leave it at that.