P2P Redd 47 - a few questions

[emrr]

In increasing gain through decrease of NFB, you may find the pentode front end loses treble response.  Be sure to watch for it. 

 

[earthsled]

Do you think I could run two pre amps from the one power supply?

The transformer would need to be dimensioned accordingly, which shouldn't be difficult considering the modest requirements of a single unit.

I'm interested in building a dual-channel unit too. What kind of transformer specs would be sufficient for this? Also, would it be wise to double the filtering and regulation components, or is this necessary?

REDD.47 power consumption...
380V @ 30mA
6.3V @ 500mA

Allied 227-0081 specs...
540V @ 120mA
6.3V @ 3.5A

Great post - thanks!

 

[guavatone]

Play around with PSU designer. Design the PSU and test with 30mA load and then 60mA.  You may want to roughly half the RC filter resistors and add another stage.

 

[earthsled]

I suppose it would make since to keep the filtering and regulation in tact for each channel as this is how a "console" setup would be. Perhaps it's just the transformer and rectifiers that would need adjustment.

2 channels of REDD.47...
380V @ 60mA
6.3V @ 1A

I wish PSU designer had a mac version. I found this: http://www.hammondmfg.com/pdf/5c007.pdf but all the two-diode full wave examples make use of the CT. Does anybody have guidelines for calculating the transformer and rectifier requirements for the type of supply used in the P2P drawing?

Also, I'm assuming that the PSU caps should be in the 450V to 500V range, it this correct?

Thanks!

 

[abbey road d enfer]

You don't need software to design this type of PSU.
You need 380V DC at 60mA, that means the xfmer must deliver 380Vac/1.414 under 60mAx1.414.
The smoothing caps have to be determined by how much ripple is tolerable; this depends on the circuit (SE vs. PP, NFB or not, triode vs. pentode,..). The EMI guys have already done that for you.
I would certainly double filtering and regs.

 

[earthsled]

that means the xfmer must deliver 380Vac/1.414 under 60mAx1.414.

Ah, okay, so the ideal transformer for two modules would be about 540V (270-0-270) at 85mA -- correct?

What about the diodes? 1N007 are rated for 1A each, would that mean they combine to handle 2A total?

Thanks!

 

[abbey road d enfer]

That's correct, but not the whole story. Rectifier diodes in tube gear are very often over-spec'd, because the current is pulsated. For 1 Amp of DC output, the peak current is probably more like 10-20 Amps. I wouldn't want to make a 2 Amps (well, 1.4 Amp actually) PSU with just a pair of 1N4000x.
OTOH, 1 Amp is the continuous rating but they are spec'd for 10-30A peak.
For 60mA, they'll be fine.

 

[earthsled]

Hi all,

I've been comparing the original EMI schematic, the Gorbutt schematic, and the P2P layouts. For clarity, I decided to add part numbers to the documents - hopefully this will be useful to the DIYers out there!

I also found some discrepancies...
1. R22 from the EMI schem seems to have been omitted from the Gorbutt version -- is it needed?
2. R19 and R20 of the "alternative circuit" seem to have shifted to the left on the Gorbutt schem. -- I'm guessing this is irrelevant since there's probably no one using the REDD.43/D23/2 power unit.

 

[earthsled]

...here's the P2P layout

The Zobel circuit has been mentioned, but I'm wondering about the extra capacitance in parallel with C1. What's the purpose of this?

Also, note the correction between R12 and C6.

 

[earthsled]

...and finaly the P2P power supply section

no questions here, just helpful part numbers! 

please let me know if I missed something, or if there are corrections.

Thanks!

 

[bernbrue]

In increasing gain through decrease of NFB, you may find the pentode front end loses treble response.  Be sure to watch for it.

Hi Emrr,
in my prototype I don´t want to use the NFB at all. Is there a way to compensate the treble roll-off? Since I don´t need the huge amount of gain from the EF86, wiring it as triode could be a solution. Is there any other alternative? Sorry for being a bit off topic. Go on earthsled, good work!
regards
Bernd

 

[abbey road d enfer]

1. R22 from the EMI schem seems to have been omitted from the Gorbutt version -- is it needed?

This res is not mandatory. It is just there to make sure there is a discharge path for B+ when turning off the unit. Just there to avoid jolts.

 

[abbey road d enfer]

I'm wondering about the extra capacitance in parallel with C1. What's the purpose of this?

Many people believe that putting a small value film cap in parallels with a 'lytic will cure some imaginary ill caused by ESR (Equivalent Series Resistance).
You can experiment with and without it and make your choice.

 

[earthsled]

...here's my attempt at reading the original minimum values from the EMI schematic.
The items in red are my best guess. Please let me know if you have corrections or more info.

I wonder if "H.S." type resistors are like these: http://www.willow.co.uk/HS.pdf

Enjoy!

 

[abbey road d enfer]

I don't know what H.S. means, but I don't think it's related to a particular brand. These HS branded res are high-voltage types with standard values ranging from 1000k to 1G.
The high-voltage condition could be pertinent for R2, 3, 7, 8, 9...but not for R1, 4, 5, 6, 10, 12...

 

[earthsled]

Perhaps H.S. means "high stability"? In any case, modern metal-film resistors seem to be an appropriate replacement. I'm planning to use 1/2W resistors for everything except for R11 and R18 to 21 just to keep it simple. Is there any advantage to keeping the carbon comp resisters where specified?

Speaking of resistors, has anyone in the group worked up a solution for a gain control? It looks like the Revolution R.47 MkII used an 11-position 20dB stepped attenuator before the input transformer. This seems practical. Is it the best way?

Also, there are some who say the folks at Abbey Road would patch two REDD.47 amplifiers in series when more gain was needed. Is there anything technically wrong with this idea?

Thanks!


Attached - REDD.47 minimum values v2 -- corrected "C24" to read "C12A Etc."

 

[earthsled]

...here's version 2 of my schematic edit. This one has part numbers and cap ratings!

 

[abbey road d enfer]

Perhaps H.S. means "high stability"?

Quite possible...

In any case, modern metal-film resistors seem to be an appropriate replacement. I'm planning to use 1/2W resistors for everything except for R11 and R18 to 21 just to keep it simple.

That's what I would do...

Is there any advantage to keeping the carbon comp resisters where specified?

No. Some would argue that carbon comp res give a certain flavour to the sound (mainly crackle...), but these are used in positions where they could hardly have any serious impact on sound.

Speaking of resistors, has anyone in the group worked up a solution for a gain control? It looks like the Revolution R.47 MkII used an 11-position 20dB stepped attenuator before the input transformer. This seems practical. Is it the best way?

Why not a good old log pot? And there is certainly a possibility to reduce gain by increasing NFB; this has not been implemented by the EMI engineers probably because there was no need for less than 34dB gain at the time, but today I find myself quite often using only 25-30dB gain on very close-mic'd loud sources.

Also, there are some who say the folks at Abbey Road would patch two REDD.47 amplifiers in series when more gain was needed. Is there anything technically wrong with this idea?

That was the standard procedure in many studios around the world, considering that most EQ's were passive, so using EQ on an already feeble signal would make adding an amp a necessity. Inserting an additional attenuator for trimming the operating level was also customary. Although it seems a pity to amplify then attenuate and amplify again, if done with some technical background and a pinch of common sense, it works adequately.

 

[earthsled]

Why not a good old log pot? And there is certainly a possibility to reduce gain by increasing NFB; this has not been implemented by the EMI engineers probably because there was no need for less than 34dB gain at the time, but today I find myself quite often using only 25-30dB gain on very close-mic'd loud sources.

A log pot is a fine idea. Would you suggest adding one between the input transformer and the grid of V1? What value would be best?

On second thought, the original stepped-gain w/ fine-gain config makes sense for stereo applications. This could benefit a two channel build.

From what I understand, NFB (net feedback) is controlled by the 3-position gain switch. Are you suggesting that the switch could be modified to have additional steps for lower gain settings? Is there any disadvantage to this approach?

Thanks!

 

[earthsled]

Reading thought this post http://www.groupdiy.com/index.php?topic=7292.60 (reply #76) indicates the original A92 input transformer did NOT have a center tap on the secondary. The EMI schematic shows C13 and R1 tied to pin 6 of the transformer which apparently was an unused pin where the components were mounted.

In the attached schematic I have removed the center tap on the secondary of T1, added R22, and rearranged R19 and R20 in the "alternative" supply circuit to match the original EMI specs.

Cheers!