Push Pull ECL86/6GW8
Hammond AO-44 Reverb Amp to Hi-Fi Amp Conversion
Hammond schematics here and elsewhere on the Net
I just "got" one of these reverb organ amps. I wanted to turn it into a small mono desk amp but it seems there are several ways of doing it. Either with an input tube (12AX7) or the other designs without. My preamps put out 2rms-3rms depending on which model im using. So would would a need a gain stage or can i just put the signal right into pin 1 of the v1 ecl86? and since this will be an amp and not integrated, Do i just do a small input cap and then 1meg to ground to bypass the pot?
Hammond AO-44 Reverb Amp to Hi-Fi Amp Conversion
Hammond schematics here and elsewhere on the Net
I just "got" one of these reverb organ amps. I wanted to turn it into a small mono desk amp but it seems there are several ways of doing it. Either with an input tube (12AX7) or the other designs without. My preamps put out 2rms-3rms depending on which model im using. So would would a need a gain stage or can i just put the signal right into pin 1 of the v1 ecl86? and since this will be an amp and not integrated, Do i just do a small input cap and then 1meg to ground to bypass the pot?
You don't need any additional gain, when using a preamp. If anything, you will have a "hair trigger" volume control.
IMO, it's easy enough to retain Hammond's design. A few tweaks are (IMO) indicated.
Discard everything to left of C5. R7 and R11 become 100 KOhms/metal film. C5 becomes 56 nF. R9 stays at 220 KOhms, but inductive wirewound construction is in order. C6 becomes 1 μF. C7 and C8 become 100 nF. The 100 μF. bias network 'lytic increases to 270 μF. Place a 22 pF. mica or NPO ceramic cap. across the O/P trafo secondary.
Oh yeah, add 100 Ω stoppers at the triode grids and 1 KOhm stoppers at the pentode control grids. Carbon composition is the best construction type for grid stopper service.
IMO, it's easy enough to retain Hammond's design. A few tweaks are (IMO) indicated.
Discard everything to left of C5. R7 and R11 become 100 KOhms/metal film. C5 becomes 56 nF. R9 stays at 220 KOhms, but inductive wirewound construction is in order. C6 becomes 1 μF. C7 and C8 become 100 nF. The 100 μF. bias network 'lytic increases to 270 μF. Place a 22 pF. mica or NPO ceramic cap. across the O/P trafo secondary.
Oh yeah, add 100 Ω stoppers at the triode grids and 1 KOhm stoppers at the pentode control grids. Carbon composition is the best construction type for grid stopper service.
Ok i dont know what a stopper is other then its used instead of feedback fro high freq oscillation. i also dont know were it "goes".
triode grids are pin 1 and pentode grids are pin 8. Then you got the suppressor grid pin 3 that connects to the cathode.
About c5. Does it have to be 56nf? I have some extra black gate 22uF with some wima 1uf bypass caps left over from my dac build. Or could i use a vitaminQ 180nf from beezar.com. Or if you have other recommendations id love to hear it. Input caps are sort of an art.
Also is it a good idea to pop in a .1uf film/box cap across the 270uf?
And finally, How come i only see SE amps with a choke in the powersupply? Is it not needed in most PP desgins?
triode grids are pin 1 and pentode grids are pin 8. Then you got the suppressor grid pin 3 that connects to the cathode.
About c5. Does it have to be 56nf? I have some extra black gate 22uF with some wima 1uf bypass caps left over from my dac build. Or could i use a vitaminQ 180nf from beezar.com. Or if you have other recommendations id love to hear it. Input caps are sort of an art.
Also is it a good idea to pop in a .1uf film/box cap across the 270uf?
And finally, How come i only see SE amps with a choke in the powersupply? Is it not needed in most PP desgins?
Last edited:
A stopper is installed to prevent parasitic oscillation from occurring. Mount the resistor's body as close to the socket's solder lug as can be.
That 56 nF. part is carefully computed. It sets the LF 3 dB. down point of the amp at roughly 28 Hz. The lowest note a double bass makes is 31 Hz.
Rolling the deepest bass tones, along with infrasonic noise, off protects the O/P trafo against core saturation, due to a GNFB generated LF error correction signal.
High quality 'lytics, like those offered by Panasonic and Nichicon, will get the job done, with a bypass. If you are fanatic, bypass the 270 μF. part with a 22 μF. MultiCap. A 100 nF. part's effect is vanishingly small. IMO, the expense of a proper bypass part is not warranted.
That 56 nF. part is carefully computed. It sets the LF 3 dB. down point of the amp at roughly 28 Hz. The lowest note a double bass makes is 31 Hz.
Rolling the deepest bass tones, along with infrasonic noise, off protects the O/P trafo against core saturation, due to a GNFB generated LF error correction signal.High quality 'lytics, like those offered by Panasonic and Nichicon, will get the job done, with a bypass. If you are fanatic, bypass the 270 μF. part with a 22 μF. MultiCap. A 100 nF. part's effect is vanishingly small. IMO, the expense of a proper bypass part is not warranted.
Do you have the math for the input cap? Im planning on using this for a small desk amp thats no realy going to be playing below 80hz with the bookshelf. It might be nice to use that for a highpass filter. I dont know what the op trans is capable of playing down to.
So i just put the stoppers inbetween:
r7 and pin 1
r11 and pin1
r14 and pin8
r15 and pin8
as close to the tube as possible?
Also what type/brand/place to buy do you recomend for the inputcap? It looks like all my normal dac output caps wont do.
I guess i need to read up on my cap coupling. all ive used are dc strait amps with dacs output like this:
The γ1 Modular Miniature DAC
The γ2 Compact High Performance DAC
or a charge pump output that doesnt need a cap:
GrubDAC Schematic
Guess some amps still need inputcaps even if the input isnt lifted above ground.
So i just put the stoppers inbetween:
r7 and pin 1
r11 and pin1
r14 and pin8
r15 and pin8
as close to the tube as possible?
Also what type/brand/place to buy do you recomend for the inputcap? It looks like all my normal dac output caps wont do.
I guess i need to read up on my cap coupling. all ive used are dc strait amps with dacs output like this:
The γ1 Modular Miniature DAC
The γ2 Compact High Performance DAC
or a charge pump output that doesnt need a cap:
GrubDAC Schematic
Guess some amps still need inputcaps even if the input isnt lifted above ground.
Correct.
The law (formula) for RC filters follows immediately.
(2) (Π) (FHertz) (CFarads) ROhms = 1
If you make the I/P cap. 0.047 μF., a very common value, all should be well. This is a low voltage situation. Remember, a "standard" CDP puts out 2 VRMS. So, the stuff you use in and around DACs could easily be suitable. A Soviet surplus paper in oil (PIO) part would be terrific here.
Either Panasonic ECQ-P(U) series or 716P series Orange Drops, in 400 WVDC, will do nicely for C7 and C8. Those varieties are modest cost Aluminum foil/polypropylene film, with Copper leads. Great bang for the buck.
For C6, a Solen metalized polypropylene (MPP) part (also 400 WVDC) bypassed by the same part used for C7 and C8 will be cost effective and good performing.
Contact Jim McShane. His prices, including shipping, are fair and the man KNOWS his stuff. Link this thread in the EMail you send, to ensure things turn out properly.
Well Jim got back to me.
He doesnt know of any inductive wire wounds. Should i just get a normal ww from mouser?
He doesnt have any 1uf caps for C6, only .68uf
Hes asking what the construction? of the electrolytic should be (maybe hes asking low ohm or something else, panasonic FC or something). I assume it should also be 50v, but Forest is using a 470v unit even though he lists 50v on his page.
He says that a 22uf multi cap will be huge for the bypass. Should i not bother with it or get something smaller?
He doesnt know of any inductive wire wounds. Should i just get a normal ww from mouser?
He doesnt have any 1uf caps for C6, only .68uf
Hes asking what the construction? of the electrolytic should be (maybe hes asking low ohm or something else, panasonic FC or something). I assume it should also be 50v, but Forest is using a 470v unit even though he lists 50v on his page.
He says that a 22uf multi cap will be huge for the bypass. Should i not bother with it or get something smaller?
Last edited:
Unless specifically stated otherwise, wirewound resistors are inductive. You pay extra for non-inductive construction. For R9, a series combination of Mouser stock # 71-CW010100K0JE12 and Mouser stock # 71-RN65D-F-121K will do nicely.
Mouser stock # 871-B32653A4105J, plus a film bypass, will work well as C6.
The schematic shows 5 V. at the pentode cathodes. A 25 WVDC part is quite sufficient. Obviously, 50 WVDC is safe. A Nichicon KZ, without any MultiCap, is FINE. A film bypass here is for people with OCD! A 270 μF. part may not be available, but 330 μF. is and it's perfectly satisfactory. I know for a fact that Nichicon makes a KZ series 330 μF./25 WVDC part. Whether or not he stocks it, only Jim can say. This part needs to be low impedance/high bandwidth. Cheap Chinese guano need not apply for this position.
BTW, it likely that the OEM resistors have drifted out of spec., over time. Resistors that have drifted should be replaced. With exception of R17 and the parts I mentioned early on, 0.5 W./5% Xicon Carbon film parts from Mouser are suitable. Use a wirewound part for R17. Note that at least a 3 W. rating is necessary.
For a 220K resistor there will be a LOT of spiraling required to get that value, so the inductive component won't be much different than if a wirewound of that value could be found. I think a CF or MF part would do well - what wattage is needed?
Back ordered, over 10 weeks wait is likely before the part could be received.
An aside - right now with the recession the number of back orders is through the roof. You can't count on anything being in stock anymore - even if it's items that have been popular in the past. And Mouser is actually one of the better stocked distributors.
As well, with 100K here why do we need a 1 uf coupling cap? 1 uf and 100K give a corner frequency of about 1.6 Hz. A .68 uf would only raise that to 2.3 Hz or so. A .47 uf would give a corner of 3.4-ish Hz. So with even a .47 uf we get a corner frequency that is almost 3 octaves below 20 Hz. Maybe I am missing something this early in the AM, but I can't figure why a 1 uf cap is needed. Maybe I looked at the wrong schematic.
Anyway, I'll be happy to help any way I can!
Well a few more of these have started to show up on ebay. These seem like bargains (aka you get the power transformer for $50 and a free amp) and i was going to snatch a few up for mono blocks for the house. However my house speakers are all 4ohm. I think, but im not possitive that the output transformers are 5k.
Can i run 4 ohm speakers with a 8ohm output transformer as long as i keep the output below the 8ohm wattage limit? Or is it just gona sound bad.
If i replace the output transformer with an edcor, do i need a PP series or do i do a SE series?( i know you cant run a se from a pp trans but im not up on that tech yet)
This one caught my eye due to the multi tap:
EDCOR - CXPP30-MS-5K
Can i run 4 ohm speakers with a 8ohm output transformer as long as i keep the output below the 8ohm wattage limit? Or is it just gona sound bad.
If i replace the output transformer with an edcor, do i need a PP series or do i do a SE series?( i know you cant run a se from a pp trans but im not up on that tech yet)
This one caught my eye due to the multi tap:
EDCOR - CXPP30-MS-5K
I spoke to Jim and we kicked the why and where for around. A parallel combination of a 680 nF. MMP and a 470 nF. ECQ-P(U) will work well as C6. There's no need to put up with that blasted 10 week delay.
The reason the high pass pole formed by C6 and R11 is set so close to zero Hz. is to protect against the possibility of phase shift oscillation. The phase shifts in the O/P trafo are an unknown quantity. Aside from C6/R11, there is a high pass pole at each pentode's control grid. The shared pentode bias network is a high pass pole too. At F3 in a RC filter, the phase shift is 90o. 4X 90 = 360, which would turn the feedback loop (R16/R8) from negative to positive and destabilize the circuit. Spreading the pole "corner" frequencies out dodges the bullet. If you want more info., Google Barkhausen oscillation criterion.
The reason the high pass pole formed by C6 and R11 is set so close to zero Hz. is to protect against the possibility of phase shift oscillation. The phase shifts in the O/P trafo are an unknown quantity. Aside from C6/R11, there is a high pass pole at each pentode's control grid. The shared pentode bias network is a high pass pole too. At F3 in a RC filter, the phase shift is 90o. 4X 90 = 360, which would turn the feedback loop (R16/R8) from negative to positive and destabilize the circuit. Spreading the pole "corner" frequencies out dodges the bullet. If you want more info., Google Barkhausen oscillation criterion.
Well im getting my mouser BOM together
They have the orange drops in stock
the KZ comes in 220uf and 330uf, not 270uf. I assume i go with the larger.
with the pentode/cathodes only being 5v, you think i can use my favority headphone supply cap, or is this in the audio path a low ogm caps like the pan fc's should be used?
Only prob is that they are 16v and the smallest is 1200 uF
http://www.mouser.com/ProductDetail...=sGAEpiMZZMtZ1n0r9vR22cURos1IcPUppthjTkR22II=
647-UHZ1C122MPM
R15 goes across the primaries with a 1nf cap. the 350v units at mouser are 45cents, the mill spec 500v ones are $2.60.
Now i just have to work on some oil in paper input caps and that darn 1uf c6...
The unit does have 1 1mdf cap that is huge (muliticap style) rated at 100vcd that says gudeman on it. (C4)
They have the orange drops in stock
the KZ comes in 220uf and 330uf, not 270uf. I assume i go with the larger.
with the pentode/cathodes only being 5v, you think i can use my favority headphone supply cap, or is this in the audio path a low ogm caps like the pan fc's should be used?
Only prob is that they are 16v and the smallest is 1200 uF
http://www.mouser.com/ProductDetail...=sGAEpiMZZMtZ1n0r9vR22cURos1IcPUppthjTkR22II=
647-UHZ1C122MPM
R15 goes across the primaries with a 1nf cap. the 350v units at mouser are 45cents, the mill spec 500v ones are $2.60.
Now i just have to work on some oil in paper input caps and that darn 1uf c6...
The unit does have 1 1mdf cap that is huge (muliticap style) rated at 100vcd that says gudeman on it. (C4)
Last edited:
The 716P "Orange Drops" with copper leads are preferred to the 715P "Orange Drops" that Mouser stocks because the 715Ps have steel leads.
Mouser stocks 715P series Orange Drops, with yucky steel leads. You want 716P Series Orange Drops, whose leads are Copper. Get back to McShane for all of the PP film parts. Check around and you'll find that Jim's prices are fair. Jim may be able to help with the I/P PIO too.
Yes, use a 330 μF. KZ in the cathode bypass role.
Remove the Gudeman part carefully. It may be PIO. If it's not leaking, both physically and electrically, it could find its way into another project.
This post added due to edit time limit expiring.
Hammond's schematic shows a 2000 WVDC part in the network across the O/P trafo primary. They MEAN it!
That network is present to suppress ringing in O/P trafo. Some very nasty inductive kick back spikes are present. If anything, increase the WVDC in that position. If the OEM part looks like a small domino, it's mica and you should leave it be.
Hammond's schematic shows a 2000 WVDC part in the network across the O/P trafo primary. They MEAN it!
That network is present to suppress ringing in O/P trafo. Some very nasty inductive kick back spikes are present. If anything, increase the WVDC in that position. If the OEM part looks like a small domino, it's mica and you should leave it be.What about leakage, DCR, etc.? Mere retention of nominal capacitance is (IMO) hardly sufficient. Also, the sonics of modern 'lytics beat the living guano out of vintage stuff.
Well should i just replace it with a standard $4 450v 100uf one from mouser and pop into C11's spot? Or am i going to pop the rec tube so im better off with multi small ones?
UPW2W101MRD Nichicon Aluminum Electrolytic Capacitors - Leaded
UPW2W101MRD Nichicon Aluminum Electrolytic Capacitors - Leaded
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.