the Poor Man 660 support thread

[jackies]

:
Now there is solid state sidechain boost circuit, why not go a step further and make a completely solid state sidechain?
A chipamp or a chip driver with a couple output transistors?
High power is easily attainable, effectively halving the number of tubes and transformers to deal with, as well as reducing psu load...

 

[lolo-m]

:
Now there is solid state sidechain boost circuit, why not go a step further and make a completely solid state sidechain?
A chipamp or a chip driver with a couple output transistors?
High power is easily attainable, effectively halving the number of tubes and transformers to deal with, as well as reducing psu load...

It's easy to do it but you won't have the distorsion of the tube sidechain amp...

 

[radiance]

:
Now there is solid state sidechain boost circuit, why not go a step further and make a completely solid state sidechain?
A chipamp or a chip driver with a couple output transistors?
High power is easily attainable, effectively halving the number of tubes and transformers to deal with, as well as reducing psu load...

I like the idea. I'm not good at tube circuits and maybe the 5687's bring in some character to this compressor but AFAIK all that needs to be done is replace R7/8 by a buffer opamp, followed by power amp (IC's , transistors whatever) which feeds the 1n4148 bridge. Is it really this easy?
I'm not really familiar with power amps ...how many watt's should be OK?

 

[radiance]

:
Now there is solid state sidechain boost circuit, why not go a step further and make a completely solid state sidechain?
A chipamp or a chip driver with a couple output transistors?
High power is easily attainable, effectively halving the number of tubes and transformers to deal with, as well as reducing psu load...

It's easy to do it but you won't have the distorsion of the tube sidechain amp...

Is this something we want? Is this what makes this comp special?

 

[Kingston]

:
Now there is solid state sidechain boost circuit, why not go a step further and make a completely solid state sidechain?
A chipamp or a chip driver with a couple output transistors?
High power is easily attainable, effectively halving the number of tubes and transformers to deal with, as well as reducing psu load...

It's easy to do it but you won't have the distorsion of the tube sidechain amp...

yes but that is probably not a problem and possibly could not be heard. The poorman 5687 sidechain is pretty much linear for example.

the vari-mu sound comes from the grid curves and time constant networks. The sidechain amp itself is "just" a very linear booster.

another topic would be if the sidechain amp wasn't linear and what interesting effects could be achieved with this. Clearly not everything has been done on the vari-mu front just yet.

 

[Kingston]

:
Now there is solid state sidechain boost circuit, why not go a step further and make a completely solid state sidechain?
A chipamp or a chip driver with a couple output transistors?
High power is easily attainable, effectively halving the number of tubes and transformers to deal with, as well as reducing psu load...

It's easy to do it but you won't have the distorsion of the tube sidechain amp...

Is this something we want? Is this what makes this comp special?

No, and no.

I think the ultimate vari-mu compressor could be made much better with a chipamp sidechain (transformers or not). We can have much more linear power (and high voltage) output, thus resulting in faster attack times a well. Easier on the PSU, because we would only have the vari-mu tubes, and no more inefficient sidechain tube amps.

Sound would still be completely "tube". Only control signal touches the chip, which will then go through caps (forming the time constants), thus eliminating all evidence of "cheating".

The topic has reared its head every once in a while. Not sure if anyone tried it yet.

Wish I had the skills (or maybe just the courage) to try this. Sounds much more exciting than yet another fairchild clone.

 

[lolo-m]

I think the ultimate vari-mu compressor could be made much better with a chipamp sidechain (transformers or not). We can have much more linear power (and high voltage) output, thus resulting in faster attack times a well. Easier on the PSU, because we would only have the vari-mu tubes, and no more inefficient sidechain tube amps.

Sound would still be completely "tube". Only control signal touches the chip, which will then go through caps (forming the time constants), thus eliminating all evidence of "cheating".

The topic has reared its head every once in a while. Not sure if anyone tried it yet.

Wish I had the skills (or maybe just the courage) to try this. Sounds much more exciting than yet another fairchild clone.

PRR did it with is varimu comp...
2 of us are working on that hybrid varimu concept. Easy or not, depends of the choosen varimu tube.

 

[ChuckD]

Although I did not measure the noise floor when I first completed this build I think this EMI problem only just started happening after I was using it all the time.
The reason I mounted the EDCOR on the outside was the reports of heat issues and I am thinking this transformer cannot handle the current draw required and started the EMI problem due to the large current draw and heat radiated???

Could this be true ? If so everyone has to replace theirs.
Anyone have some insight on my issue?

I have ordered a Rondo from Volker so I think mine will be solved with that.

Also ! I made a mistake before my UTC A15s are mounted at T1 and T3.  Edcors at T2 T4.
The EMI interferences enters at T1 since it increases with gain.

As for the comments on the Solid state SC? Works for TubeTech. People rave about their CLB1 which is basically Analag's Aopto with a solid state side chain.



Chuck

 

[ChuckD]

Also still waiting on a response from someone further back asking about how to match tubes.
Without a tester....

I can;t remember who asked but someone did.

Chuck

 

[radiance]

I have ordered a Rondo from Volker so I think mine will be solved with that.

AFAIK these have a 6A 9V heater secondary, so underspecked, same as Edcor.
I have one of thes as well but I only power one 660 with it. The other heater has his own 9V 8A toroid. This way both transformers get not even mildly warm. Slow start heater regulators get really hot though.These need serious heatsinking.

 

[Kingston]

The reason I mounted the EDCOR on the outside was the reports of heat issues and I am thinking this transformer cannot handle the current draw required and started the EMI problem due to the large current draw and heat radiated???

Could this be true ? If so everyone has to replace theirs.
Anyone have some insight on my issue?

Yes the large current draw and the transformer being used close to it's specified (and physical!) maximum can result to your situation, although I would first point at wiring and shielding and think about a redesign of those.

But I'm guessing your transformer is physically "singing" as well?

This is why I specified my projects custom transformer to 15V 4A for two channels of heaters running in series (to save on regulation). That means about 9V 8A for your normal parallel heaters (but now two regulators needed). This was also an order from a highly trusted transformer manufacturer, which means the transformer is more than likely something like 15V 5-6A in reality (again, 9V 10A!). More than ample headroom. But that's how you avoid problems with heating, underpowering and EMI. The golden rule is spec your transformer current so you'll only ever run it at about 60% from its rated max. Add even more headroom if you have any doubts about the manufacturer as they all have different concepts of "max".

 

[ChuckD]

Thanks for the info!

I have ruled out wiring since as I said, by simply creating some large shield for both the  transformers outside the case  (The EDCOR Power and the UTC A15s)  I immediately reduced the 60Hz and 180Hz from -55db to -67db.

The Edcor power gets really hot which implies it is working too hard I think.
So you're right  it is under spec'd as you mention.

Nothing seems to be physically "singing" though.


Chuck

 

[Kingston]

I have ruled out wiring since as I said, by simply creating some large shield for both the  transformers outside the case  (The EDCOR Power and the UTC A15s)  I immediately reduced the 60Hz and 180Hz from -55db to -67db.

That's interesting. I have several projects that might benefit from such shields. I seem to recall you mentioning grain oriented steel sheets somewhere. Where did you get some, and how did you end up installing them?

I'm always up for more shielding lore.

 

[lolo-m]

Also still waiting on a response from someone further back asking about how to match tubes.
Without a tester....

I can;t remember who asked but someone did.

Chuck

If someone need a simple to do tube match/tester for the 6BC8 in PM670 circuit, I've got a file ready to be emailed. Sorry, as my HD crashed I lost the sofware to upload it so if someone can do it for me, PM me and I'll email him the file as soon as possible as I'm really busy at the moment.
It's a really simple concept. When you'll see it you'll understand ! 2 resistors, 2 pots, 1 cap and one vumeter (the one you use in your PM), PM670's PSU.
You'll have to test each triode one by one...

 

[Kingston]

I lost the sofware to upload it so if someone can do it for me, PM me and I'll email him the file as soon as possible as I'm really busy at the moment.

You could attach an image or text file to your forum post, as long as it's not too big.

 

[radiance]

...or use the group diy gmail account for files up till 15mB..(or whatever it is gmail allows you to attach)

 

[lolo-m]

Ok, I never checked the additional options  :-[ !

To supply the filament you can set the +17V rail of the PSU660 to 6,3V, it'll work to test one tube without any problem.

The principe of the match is : plate current is fonstion of grid tension. If the plate current of one of the triode is close to plate current of the other triode at several test points (let's say 6, it's enough), your tubes are matched.

Procedure :
not to damage your tube never set the grid pot at more than -2V
1°) set the pot to -2V and adjust the vu pot to read 0dB.
2°) set the pot to -4V, and note on and excell sheet the value of the vu
3°) set the pot to -6V, and note the value... and so on...
4°) change the triode connection, and start the second triode test from -2V without readjusting the 0Vu reading

I talked about an excell sheet as it's easy to trace curves with it and it's really easy to see which triodes tracks well together.

You can easily add some lorlin to do the test quickier, for grid tension and cathode/grid/plate connection. A PCB is easy to do so...

 

[lolo-m]

One thing I forgot :
Use the same value for the Vu Pot than the one you used in your PM670.

Have fun !

 

[mrclunk]

Nice one Lomo, i shall give it a go! ;D

 

[lolo-m]

Nice one Lomo, i shall give it a go! ;D

Thanks  !

Another point is you may not find 4 dual triode with matched curves. In that case you've got several options :
- Buy some more 6BC8 (but it's expensive and you may not have matched triodes in the end)
- Make a quite good match adding the current (Vu readings) of the 1rst triodes of 2 different tubes, and the current of the 2nd triodes of the same 2 different tubes. It's not perfect but close enough to have quite well matched dual triodes in the circuit. The output transformer will cancel the main thump. There again excell will save you a lot of time.

But anyway it's pretty long to do those tests so be patient  !