Motown Direct Amplifier-inspired Preamp?

[ruffrecords]

Hey Ian,
this might be considered a thread hi-jack but worth a shot:

Do you have any experience with that particular Don Audio power transformer? (220V AC, 9V AC, 5Vac)
I'm trying to help a friend of a friend in mechanically quieting down an RS.124 type compressor that has a particularly mechanically noisy EI type power transformer.
Options are limited because we need more than the usual 6V3 heater due to running heaters DC with enough margin for the regulator drop out hence...

Anyway, some torroids I've tried radiated horrendous fields so am hoping this baby is good to go for the job.

Thanks mate

As you were gentlemen.

No experience of that particular transformer but I have used several toroids from Don Audio and he has even made some custom ones for me. Extraneous fields was never a big problem with them.

Cheers

Ian

 

[Winston OBoogie]

No experience of that particular transformer but I have used several toroids from Don Audio and he has even made some custom ones for me. Extraneous fields was never a big problem with them.

OK thanks Ian. I'll grab one of these then,
Cheers

 

[solkatten]

Don-Audio trafos:​

Audio Toroidal Transformer "Mila" - Sec.: 250vct, 6.3vct, 48v​

Very nice and quite, have used in 2 diffrent preamps with ac-heathers works great

Audio Toroidal Transformer - Pri.: 2x115v- Sec.: 240V, 50V, 2 x 12V​

Had to build a screenig can to silence the trafo. Maybe the the switching noise from rectified heathers made the trafo sing

 

[untune]

Heater voltage can easily be elevated without a centre tap.

The Don Audio GSSL transformer at about £40 would also do the job. Its 9V secondary can be used to make a dc heater supply for those who prefer it.

An alternative is to use a small SMPS for the heaters. Then there afre lots of small low cost HT only transformers to choose from which are available at Mouser/Digikey.

I just checked the 6V6 datasheet and its heater is also 0.45A just like the 6AQ5. So from the power supply point of view there is no disadvantage in using the 6V6 but it does give you the option of raising the CF current if you want to.

It is not a complicated circuit. Point to point would be perfectly feasible for DIY

Thanks Ian; Lots of options. I started to draw up a very rough schem with two power options, the Carnhill VTT2032 240v and the Hammond 370AX 240-0-240. Point-to-point would be my preferred build too, simple enough that tag strips attached to the valve socket fixings would be sufficient.

I'm curious about the way the NFB/gain is implemented in the Pultec variants; the MEQ5 is fixed with a 20k off the 6AQ5 cathode feeding back to the 1st 1/2 12ax7 cathode resistor of 150R forming a voltage divider. The MB-1 has a network of additional resistors and a cap; this I believe will be related to loading the input transformer (which is not necessary for the intended application here so can be omitted.) There were some comments that were made about it in this thread which were interesting.

Therefore the 82K can go; and there needs to be a grid leak of anywhere from... 470K up to 5M6 based on the discussions earlier in the thread. Does this get referenced to ground or does it directly replace the 82K which connects between the 180K/0.25u cap? I guess this would necessitate a DC blocking cap on the input. I studied your EZ tube mixer mic pre schematic and saw you have a similar arrangement with a 470K grid leak, connected to the junction of a 470 and 47K. I've attached an MB-1 and MEQ-5 schematic with a colour split for convenience.

Hey Ian,
this might be considered a thread hi-jack but worth a shot:

Do you have any experience with that particular Don Audio power transformer? (220V AC, 9V AC, 5Vac)
I'm trying to help a friend of a friend in mechanically quieting down an RS.124 type compressor that has a particularly mechanically noisy EI type power transformer.
Options are limited because we need more than the usual 6V3 heater due to running heaters DC with enough margin for the regulator drop out hence...

Anyway, some torroids I've tried radiated horrendous fields so am hoping this baby is good to go for the job.

Thanks mate

As you were gentlemen.

No worries Winston, good to know the details on the power transformer options anyway I had read reports that some of the Hammonds vibrate quite noticably but I don't know how common it is

 

[untune]

Just related to the above, redrawn to compare differences

 

[abbey road d enfer]

Just related to the above, redrawn to compare differences

The mixer pre stage won't work, unless you want a fuzz option. Bypassing the cathode for varying gain works only in common-cathode stages. Here you have a cath-foll, which has essentially unity gain, until it runs out of juice and distorts.

 

[Winston OBoogie]

Audio Toroidal Transformer - Pri.: 2x115v- Sec.: 240V, 50V, 2 x 12V​

Had to build a screenig can to silence the trafo. Maybe the the switching noise from rectified heathers made the trafo sing

Noted dude, thanks.
I get a might nervous around unknown torroids due to one bad experience of ordering 1000 of the buggers from Asia based on a prototype from them which worked just fine.
All of the shipped 1000 ended up needing mu-metal shields which added $20 to the BOM and negated any cost savings we'd hoped to get over buying a more reputable type.

I see that Don Audio's are made in Germany though so my confidence is higher.

 

[solkatten]

Noted dude, thanks.
I get a might nervous around unknown torroids due to one bad experience of ordering 1000 of the buggers from Asia based on a prototype from them which worked just fine.
All of the shipped 1000 ended up needing mu-metal shields which added $20 to the BOM and negated any cost savings we'd hoped to get over buying a more reputable type.

I see that Don Audio's are made in Germany though so my confidence is higher.

I dont know if where just my ex that were noisy, but the trafo had elctrostatic screen and screening shield and were supose to be quit...

 

[Winston OBoogie]

I started to draw up a very rough schem

I'm curious about the way the NFB/gain is implemented in the Pultec variants; the MEQ5 is fixed with a 20k off the 6AQ5 cathode feeding back to the 1st 1/2 12ax7 cathode resistor of 150R forming a voltage divider. The MB-1 has a network of additional resistors and a cap; this I believe will be related to loading the input transformer

On the MB-1, the input transformer is bootstrapped up by the 180K series resistor and the C25 (or is it C29, can't quite read it) on your schematic.

This enables you to strap various gain adjust resistors across the lower cathode leg without upsetting the bias a whole lot.
But it means there will be a voltage present on the input control grid so you'd want a DC blocking cap before your input terminating resistor of 470K - 5M or whatever you have.
Also, being bootstrapped, that input resistor will "look" like a higher input Z to your guitar than if it were simply a grid leak.

Without messing with the scheme, it may be preferable to go with the MEQ's way of doing it and have the gain be more of a "set it and forget it".
You could implement a variable feedback topology but you'd need to deviate from the Pultec as the feedback there is DC, so not something you'd want to adjust in situ if you're fond of your tweeters, or ears.

P.S. Whereabouts in Manchester are you? I'm a repatriated Mancunian myself.

 

[Winston OBoogie]

Or: use a high value R for the bootstrapped control grid, have an input blocking cap, then use the other side of the cap to establish your input Z.

That's probably what I should have come up with in the first place...

 

[untune]

On the MB-1, the input transformer is bootstrapped up by the 180K series resistor and the C25 (or is it C29, can't quite read it) on your schematic.

This enables you to strap various gain adjust resistors across the lower cathode leg without upsetting the bias a whole lot.
But it means there will be a voltage present on the input control grid so you'd want a DC blocking cap before your input terminating resistor of 470K - 5M or whatever you have.
Also, being bootstrapped, that input resistor will "look" like a higher input Z to your guitar than if it were simply a grid leak.

Without messing with the scheme, it may be preferable to go with the MEQ's way of doing it and have the gain be more of a "set it and forget it".
You could implement a variable feedback topology but you'd need to deviate from the Pultec as the feedback there is DC, so not something you'd want to adjust in situ if you're fond of your tweeters, or ears.

P.S. Whereabouts in Manchester are you? I'm a repatriated Mancunian myself.

Thanks for the explanation, I think it's making sense I think that cap is 0.25u, my furious scribbling to blame!

I'm inclined to agree with the MEQ 'set and forget' aproach too as that seems to be the way it was originally intended; let the instrument vol pot control the level.. I'll probably experiment with various grid leak values to see what (if any) effect on the overall tone there might be, though I'm guessing not a great deal if the bootstrapping increases input Z. I was just curious if that network was specifically related to the transformer, whereas I'm not using mic input so that can be scratched; something similar in the Gyraf G9 I think which uses a switched hi-z input jack. I might look at the stepped gain approach in the feedback at some point after trying it fixed, DC switch popping isn't my kind of groove

And small world! From Bolton originally, I'm around Worsley now so technically Salford, how about you?

 

[Winston OBoogie]

I'm inclined to agree with the MEQ 'set and forget' aproach too as that seems to be the way it was originally intended;

If you do it that way, then the feedback also helps establish the DC bias of the first stage. The ratio of the 20K series, and 150R shunt, besides determining the voltage gain, is also pulling up the cathode bias voltage so that the 150R behaves more like a capacitor bypassed resistor, or led, or diode etc...
With that scheme, the input grid isn't bootstrapped and can be simply ground referenced, so you just need your input grid to ground resistor of choice.
No need for a blocking cap.

And small world! From Bolton originally, I'm around Worsley now so technically Salford, how about you?

If I yelled or farted loudly enough, there's a chance you might hear me!

I'm technically Salford too, in Swinton. I lived in Worsley as a kid, it's a nice area. I'd prefer to be there now but, alas, bank balance forbids.
My cousin is on Mill Brow in Worsley and I go over quite a bit, was there a week ago for a stroll along the canal.

Small world for sure

 

[olhsson]

We probably have different understandings on this point or a big difference between both sides of the ocean.
In Europe in the 60's the majority of lead vocals were overdubbed, first in mono, implying at least one more generation for the backing track, then when 2-track machines came, background was on one channel and vocals on the other, subsequently mixed to a mono machiine. later there were 3-trackmachines, that actually didn't change the MO, but offered more overdub possibilities. Then it progressed into 4 track, 8 track...

In the U.S., major labels needed to pay the session musicians for all time spent overdubbing prior to the late '60s. Only the most elaborately overdubbed projects ever got written up in magazines, but it seems they were by no means SOP. I was very surprised to learn this from some older engineers because I had assumed everybody did it like we did at Motown.

 

[abbey road d enfer]

In the U.S., major labels needed to pay the session musicians for all time spent overdubbing prior to the late '60s.

It was exactly the same in Europe. Since singers were the ones that created the need for multiple takes (not everybodyis Frank Sinatra), the vocal sessions were always done after the backing track.

 

[Audio1Man]

Hi All

I have known John for many years. John gave a presentation for the Hollywood Sapphire Group April 6 about Motown and is going to do a continuation in the next few months. He lives in Culver City CA. and I won’t pass out his information unless he tells me it’s OK. We saved his presentation (Zoom) and it is not public, only for Sapphire members.

Duke

 

[ruffrecords]

Just related to the above, redrawn to compare differences

Both are equivalent, just slightly different ways of achieving the same thing. In both versions you need to develop the right dc voltage between cathode and grid which is achieved by the 120 ohms in the MB-1 and the 470 ohms in my design. The lower resistor (1500 in the MB1 and 47K in my design) allow the overall gain to be altered by bypassing this resistor with a suitable RC circuit in order to vary the negative feedback. In both case NFB extends down todc. The MB-1 was in fact the inspiration for my design (along with the V76).

I would recommend we go fro the MB-1 version but modified for a higher input impedance and with added input capacitor. That way we can tweak the gain simply by varying one resistor.

Cheers

Ian

 

[Winston OBoogie]

The lower resistor (1500 in the MB1 and 47K in my design) allow the overall gain to be altered by bypassing this resistor with a suitable RC circuit in order to vary the negative feedback. In both case NFB extends down todc. The MB-1 was in fact the inspiration for my design (along with the V76).

I would recommend we go fro the MB-1 version but modified for a higher input impedance and with added input capacitor.

A wee disclaimer: The MB-1 was not intended to have its gain be continuously adjustable, it was a set it and forget it device. Gain adjust was by installing your choice of parallel resistor to the 1500R only so, without a suitable cap that you refer to, adjusting gain will give loud cracks.

 

[untune]

If you do it that way, then the feedback also helps establish the DC bias of the first stage. The ratio of the 20K series, and 150R shunt, besides determining the voltage gain, is also pulling up the cathode bias voltage so that the 150R behaves more like a capacitor bypassed resistor, or led, or diode etc...
With that scheme, the input grid isn't bootstrapped and can be simply ground referenced, so you just need your input grid to ground resistor of choice.
No need for a blocking cap.



If I yelled or farted loudly enough, there's a chance you might hear me!

I'm technically Salford too, in Swinton. I lived in Worsley as a kid, it's a nice area. I'd prefer to be there now but, alas, bank balance forbids.
My cousin is on Mill Brow in Worsley and I go over quite a bit, was there a week ago for a stroll along the canal.

Small world for sure

Ok I see! I was really curious about the simplicity of the meq, I took a break from valve stuff for a couple of years so I'm relearning a lot of things! I was wondering if the 150 acts as cathode resistor for v1a, and the 20k for the 6aq5, or whether the DVR for the output transformer primary is more important there. Need to brush up on a lot of maths, figure out the method for drawing load lines for a cathode follower too

Haha! And no way, literally up the road from me, I actually work down by the canal (at least, pre lockdown, still remote at the moment) so I know the area quite well

Hi All

I have known John for many years. John gave a presentation for the Hollywood Sapphire Group April 6 about Motown and is going to do a continuation in the next few months. He lives in Culver City CA. and I won’t pass out his information unless he tells me it’s OK. We saved his presentation (Zoom) and it is not public, only for Sapphire members.

Duke

Hi Duke, I got in touch with Rick Chinn about contacting John a little way back, I didn't get a follow up but, in the meantime one of our members has shared some of the info we were hoping to learn as well as Acme contributing too. As with anyone he'd be more than welcome to join the discussion, I personally found his knowledge and Bob's stories fascinating in the AES video recently and everyone here would too I'm sure!

Both are equivalent, just slightly different ways of achieving the same thing. In both versions you need to develop the right dc voltage between cathode and grid which is achieved by the 120 ohms in the MB-1 and the 470 ohms in my design. The lower resistor (1500 in the MB1 and 47K in my design) allow the overall gain to be altered by bypassing this resistor with a suitable RC circuit in order to vary the negative feedback. In both case NFB extends down todc. The MB-1 was in fact the inspiration for my design (along with the V76).

I would recommend we go fro the MB-1 version but modified for a higher input impedance and with added input capacitor. That way we can tweak the gain simply by varying one resistor.

Cheers

Ian

That sounds like a good approach, doesn't necessarily have to be continuously variable but I think having a few set gain settings could be useful to allow for variations in input. It could still be 'set and forget' for the most part but you could still choose an alternative gain if you remember to kill your monitors or do it when the power is off.

I've been curious lately about the 6bx7 that was mentioned about being in the original to drive the output transformer, been reading up and managed to find three articles from the mid 50s to early 60s where people started to use it in an audio amplifier role as opposed to the intended TV application. Always using each half of the triode in push pull though, apparently the two halves are rarely matched but paralleled it has an anode resistance of 650r, wondering if it might have been used similarly to the way the 12au7 is in the mb1

 

[Winston OBoogie]

" I was really curious about the simplicity of the meq" ... "I was wondering if the 150 acts as cathode resistor for v1a, and the 20k for the 6aq5, or whether the DVR for the output transformer primary is more important there."

Yes the 150 R is V1a's cathode R, and the 20K is predominately the 6AQ5's cathode R. DCR of the OPT has no role in the biasing.

But we also have DC feedback from the 6AQ5's cathode to V1a's cathode which will act as a partial "forced bias".

It's a nice symbiotic arrangement in my opinion

Haha! And no way, literally up the road from me, I actually work down by the canal (at least, pre lockdown, still remote at the moment) so I know the area quite well

Yep, up/down the road. A stone's throw.

doesn't necessarily have to be continuously variable but I think having a few set gain settings could be useful to allow for variations in input. It could still be 'set and forget' for the most part but you could still choose an alternative gain if you remember to kill your monitors or do it when the power is off.

If you put the +tve of a capacitor (220uF/6V3 is probably close enough for rock 'n' roll) at the junction of the 120R and 1500R of the MB-1's schematic, you could then use various value resistors from the cap's -tve to ground in order to determine the amount of bypassing that the 1500R has.
Unless I've lost the plot, the ratio of 120R and 1500R gives, approx. a 20dB boost between an unbypassed 1500R (cap -tve left open, or with some arbitrary high value) and a fully 220uF bypassed 1500R (cap -tve shorted to ground).
Then there'd be no nasty DC on your switch to cause cracks and pops.

 

[Audio1Man]

Hi untune

John Windt is on Linkedin. Windt Audio. If you send me Email [email protected] I will sent it to John.