Anything to report on your 'potato amp' project yet?

Cheers, John Stewart

In article , wrote:

Anything to report on your 'potato amp' project yet?

Cheers, John Stewart


Hi John,

I think I previously reported that I had successfully bid on some E182CCs
on eBay. After Andre Jute, my potato amp mentor, announced a week ago
that he was starting his summer sabbatical, I decided that when the tubes
arrived I would put them on the shelf until the fall when Andre returns,
and resume the project then.

The tubes arrived on Wednesday, they are really cute little buggers, I
especially like the prominent grid coolers. With the arrival of the tubes
I decided to proceed without waiting for Andre's return in the fall, but
changes in the project goals will likely result in some delay before I am
ready to begin cutting metal.

The major goal that has changed is that I now want to build a stereo
version of the amplifier, rather than the mono block test bed that I had
originally proposed. The mono test bed doesn't really make much sense for
me, I know the amplifier will work, so I would like a stereo version that
I can use for extended listening tests in my system.

This change to a stereo design creates problems with the transformers. I
have the necessary transformers for a mono block version in my junk box,
but building a stereo version creates problems with both the output and
power transformers.

The power transformer I was going to use for the mono block version has a
B+ winding rated for 50 mA DC with a half wave rectifier. This easily
meets the estimated current requirement of 40 mA for a single channel, but
the stereo version will require something on the order of 80 mA. If my
math is correct, the transformer should be good for 70 mA if I use a full
wave bridge, which I planned on using in any case, but that still leaves
me 10 mA short. I would like to use this power transformer so I suppose
there are two options. Option one would be to reduce the quiescent
current of the amplifier to 35 mA per channel, running it slightly into
class AB. The second alternative would be to go with the original 40 mA
bias current and see if the transformer can take it. The 6.3 volt heater
winding will be running at only 40% of capacity, so this will help reduce
the total heat load on the transformer somewhat, possibly allowing me to
get by with an 80 mA B+ draw. I have decided to continue with this
transformer in the stereo version and see what happens.

The main problem area is the output transformers. I had planned on using
a Hammond transformer that I already have in the mono block version. I
was unable to locate this transformer in my transformer vault, I must have
taken it out for potential use in another project, and lost track of it,
hopefully I will eventually find it. This Hammond transformer had a 10k
CT primary and a secondary with taps for 4, 8, 16 Ohms. IIRC-AIMN the
size and ratings of this transformer were similar to the current 1609 with
out the Ultra Linear taps on the primary, and without the series/parallel
secondary arrangement. The fact that the transformer differs from the
current 1609 means that it would be difficult to find a second transformer
to match it for stereo, I suppose I will have to bite the bullet and by a
new pair of 1609s, unless I stumble across something else.

Hammond does seem to offer a cheap little 10k to 4,8,16 Ohms push pull
output transformer, the 125H, but the primary inductance is only 5.6 H,
too low for my taste.

For the phase splitting transformer, my original design called for a
transformer with a 15k primary and 135k CT secondary providing a 1.5X
voltage gain from the input to each grid. While that would be the
ultimate goal I am happy with my plan to forgo the extra gain and simply
use the secondary of a "line to push pull grids" transformer as a center
tapped choke to drive the out of phase grid, leaving the primary flapping
in the breeze.

While perusing the Hammond catalog I did notice that they have an
interstage transformer with the impedance ratios I want, but I suspect
that like the cheap Hammond output transformer the low inductance is
probably too low to meet my requirements.

Ideally I would like something similar to the UTC HA-106 for the input
transformer. A pair of HA-106s sold recently on eBay for $360.00 US,
about twice what I would want to pay.

So the bottom line is that I will go with the power transformer I was
planning on using in the mono block, and the center tapped chokes for
phase splitting, leaving only the output transformers to be resolved
before I can begin cutting metal.

I still haven't decided which of the three bias schemes I will use, but I
don't have to finalize that decision until the amplifier is otherwise
complete, then I can simply solder in the components for whichever bias
scheme strikes my fancy that day.

Suggestions for a pair of small low cost output transformers are welcome,
assuming they have adequate inductance, or if someone has a transformer
that matches my existing Hammond, which they would like to sell, that
would be ideal!


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

After Andre Jute, my potato amp mentor, announced a week ago
that he was starting his summer sabbatical, I decided that when the tubes
arrived I would put them on the shelf until the fall when Andre returns,
and resume the project then.

Not to worry. Go ahead.

Mr. McCoy's ego will not permit such a sabbatical.

Peter Wieck
Wyncote, PA

x

John Byrns wrote:

In article , wrote:

Anything to report on your 'potato amp' project yet?

Cheers, John Stewart

Hi John,

I think I previously reported that I had successfully bid on some E182CCs
on eBay. After Andre Jute, my potato amp mentor, announced a week ago
that he was starting his summer sabbatical, I decided that when the tubes
arrived I would put them on the shelf until the fall when Andre returns,
and resume the project then.

The tubes arrived on Wednesday, they are really cute little buggers, I
especially like the prominent grid coolers. With the arrival of the tubes
I decided to proceed without waiting for Andre's return in the fall, but
changes in the project goals will likely result in some delay before I am
ready to begin cutting metal.

The major goal that has changed is that I now want to build a stereo
version of the amplifier, rather than the mono block test bed that I had
originally proposed. The mono test bed doesn't really make much sense for
me, I know the amplifier will work, so I would like a stereo version that
I can use for extended listening tests in my system.

This change to a stereo design creates problems with the transformers. I
have the necessary transformers for a mono block version in my junk box,
but building a stereo version creates problems with both the output and
power transformers.

The power transformer I was going to use for the mono block version has a
B+ winding rated for 50 mA DC with a half wave rectifier. This easily
meets the estimated current requirement of 40 mA for a single channel, but
the stereo version will require something on the order of 80 mA. If my
math is correct, the transformer should be good for 70 mA if I use a full
wave bridge, which I planned on using in any case, but that still leaves
me 10 mA short. I would like to use this power transformer so I suppose
there are two options. Option one would be to reduce the quiescent
current of the amplifier to 35 mA per channel, running it slightly into
class AB. The second alternative would be to go with the original 40 mA
bias current and see if the transformer can take it. The 6.3 volt heater
winding will be running at only 40% of capacity, so this will help reduce
the total heat load on the transformer somewhat, possibly allowing me to
get by with an 80 mA B+ draw. I have decided to continue with this
transformer in the stereo version and see what happens.

Go easy on the input cap after the FWB. That will tend to keep the HV secondary
winding VA's under control. And the PS regulation should still be OK since I
get the sense you would prefer to run Class A. From your description of the PT,
sounds like the primary & core will be A'OK for this application, even at your
proposed 40 ma per channel.

The main problem area is the output transformers. I had planned on using
a Hammond transformer that I already have in the mono block version. I
was unable to locate this transformer in my transformer vault, I must have
taken it out for potential use in another project, and lost track of it,
hopefully I will eventually find it. This Hammond transformer had a 10k
CT primary and a secondary with taps for 4, 8, 16 Ohms. IIRC-AIMN the
size and ratings of this transformer were similar to the current 1609 with
out the Ultra Linear taps on the primary, and without the series/parallel
secondary arrangement. The fact that the transformer differs from the
current 1609 means that it would be difficult to find a second transformer
to match it for stereo, I suppose I will have to bite the bullet and by a
new pair of 1609s, unless I stumble across something else.

Hammond does seem to offer a cheap little 10k to 4,8,16 Ohms push pull
output transformer, the 125H, but the primary inductance is only 5.6 H,
too low for my taste.

The 125H should be OK since the PP E182CC/7119 will not be able to drive it at
full power. I used two of it's big brother, the 125J in the second Twin-Coupled
Amp project. I got excellent results, far beyond what you would expect from the
Hammond published spec. Since you are driving the OPT with triodes, the OPT
primary inductance is much less important than when driving with pentodes. The
1609 almost seems like overkill, but of course would give measurably better
results. Could we here the difference? I think a lot would depend on the
speaker system used & the source material!

For the phase splitting transformer, my original design called for a
transformer with a 15k primary and 135k CT secondary providing a 1.5X
voltage gain from the input to each grid. While that would be the
ultimate goal I am happy with my plan to forgo the extra gain and simply
use the secondary of a "line to push pull grids" transformer as a center
tapped choke to drive the out of phase grid, leaving the primary flapping
in the breeze.

Lloyd Peppard of Mapletree Audio Design used one of those ITs in a PP 6AS7 amp
which was sold on his web site. See the web site at

http://hollowstate.netfirms.com

But when I took a look on the iNet yesterday the circuit & amp were gone. I
thought I had a copy of his circuit here, but after a quick look I came up
empty.

While perusing the Hammond catalog I did notice that they have an
interstage transformer with the impedance ratios I want, but I suspect
that like the cheap Hammond output transformer the low inductance is
probably too low to meet my requirements.

Ideally I would like something similar to the UTC HA-106 for the input
transformer. A pair of HA-106s sold recently on eBay for $360.00 US,
about twice what I would want to pay.

So the bottom line is that I will go with the power transformer I was
planning on using in the mono block, and the center tapped chokes for
phase splitting, leaving only the output transformers to be resolved
before I can begin cutting metal.

I still haven't decided which of the three bias schemes I will use, but I
don't have to finalize that decision until the amplifier is otherwise
complete, then I can simply solder in the components for whichever bias
scheme strikes my fancy that day.

Suggestions for a pair of small low cost output transformers are welcome,
assuming they have adequate inductance, or if someone has a transformer
that matches my existing Hammond, which they would like to sell, that
would be ideal!

If I were to do it I would go the low cost route, at least for the trial before
spending serious money on exotic, boutique parts. The performance could be a
pleasant surprise!

Cheers, John



Regards,

John Byrns

Surf my web pages at, http://users.rcn.com/jbyrns/

In article , wrote:

John Byrns wrote:

The power transformer I was going to use for the mono block version has a
B+ winding rated for 50 mA DC with a half wave rectifier. This easily
meets the estimated current requirement of 40 mA for a single channel, but
the stereo version will require something on the order of 80 mA. If my
math is correct, the transformer should be good for 70 mA if I use a full
wave bridge, which I planned on using in any case, but that still leaves
me 10 mA short. I would like to use this power transformer so I suppose
there are two options. Option one would be to reduce the quiescent
current of the amplifier to 35 mA per channel, running it slightly into
class AB. The second alternative would be to go with the original 40 mA
bias current and see if the transformer can take it. The 6.3 volt heater
winding will be running at only 40% of capacity, so this will help reduce
the total heat load on the transformer somewhat, possibly allowing me to
get by with an 80 mA B+ draw. I have decided to continue with this
transformer in the stereo version and see what happens.

Go easy on the input cap after the FWB. That will tend to keep the HV
secondary winding VA's under control. And the PS regulation should still
be OK since I get the sense you would prefer to run Class A. From your
description of the PT, sounds like the primary & core will be A'OK for this
application, even at your proposed 40 ma per channel.

Good point about the input cap, I had planned on using a 1000 uF input cap
as in my "Power Amplifier without Power Transformer", but I will revise
that down by an order of magnitude. I hope I don't have to add a choke to
make up for the reduced capacitance.

The main problem area is the output transformers. I had planned on using
a Hammond transformer that I already have in the mono block version. I
was unable to locate this transformer in my transformer vault, I must have
taken it out for potential use in another project, and lost track of it,
hopefully I will eventually find it. This Hammond transformer had a 10k
CT primary and a secondary with taps for 4, 8, 16 Ohms. IIRC-AIMN the
size and ratings of this transformer were similar to the current 1609 with
out the Ultra Linear taps on the primary, and without the series/parallel
secondary arrangement. The fact that the transformer differs from the
current 1609 means that it would be difficult to find a second transformer
to match it for stereo, I suppose I will have to bite the bullet and by a
new pair of 1609s, unless I stumble across something else.

Hammond does seem to offer a cheap little 10k to 4,8,16 Ohms push pull
output transformer, the 125H, but the primary inductance is only 5.6 H,
too low for my taste.

Oops, sorry, I seem to have misquoted the secondary impedance specs for
the 125H, which should read "2,4,8 Ohms", not "4,8,16 Ohms".

The 125H should be OK since the PP E182CC/7119 will not be able to drive
it at full power. I used two of it's big brother, the 125J in the second
Twin-Coupled Amp project. I got excellent results, far beyond what you would
expect from the Hammond published spec. Since you are driving the OPT with
triodes, the OPT primary inductance is much less important than when driving
with pentodes. The 1609 almost seems like overkill, but of course would give
measurably better results. Could we here the difference? I think a lot would
depend on the speaker system used & the source material!

I beg to differ with you about the importance of the primary inductance.
The 125H and 125J both have the same 5.6H primary inductance even though
the 125J is designed to operate at a primary impedance level only 25% to
40% that of the 125H, so the 125H has two strikes to start with. Then
consider that in your amplifier design the transformers are enclosed
within two feedback loops, the output stage cathode feedback, and the
overall negative feedback loop, both of which serve to flatten the
frequency response curve, if not the low frequency power delivery
capability.

In contrast my amplifier has no negative feedback around the transformer,
I figure that the combination of the E182CC and the 125H would yield a low
frequency -3dB point of 70 Hz, too high for even my taste. I suspect that
the difference between the 1609 and the 125H would be plain, even to my
tin ears. The 125H is definitely underkill!

For the phase splitting transformer, my original design called for a
transformer with a 15k primary and 135k CT secondary providing a 1.5X
voltage gain from the input to each grid. While that would be the
ultimate goal I am happy with my plan to forgo the extra gain and simply
use the secondary of a "line to push pull grids" transformer as a center
tapped choke to drive the out of phase grid, leaving the primary flapping
in the breeze.

Lloyd Peppard of Mapletree Audio Design used one of those ITs in a PP 6AS7 amp
which was sold on his web site.

One of which "ITs"?

Suggestions for a pair of small low cost output transformers are welcome,
assuming they have adequate inductance, or if someone has a transformer
that matches my existing Hammond, which they would like to sell, that
would be ideal!

If I were to do it I would go the low cost route, at least for the trial
before spending serious money on exotic, boutique parts. The performance
could be a pleasant surprise!

The Hammond 1609s are hardly "boutique parts" and aren't that much more
expensive than the 125Hs. I like the smaller size of the 125Hs, and lower
cost is always desirable, but the 125Hs appear to come with too great a
performance compromise for this amplifier. Also my 96 dB sensitivity
speakers have "16 Ohm" voice coils, the 1609s have a 16 Ohm output tap to
match, while the 125Hs do not.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

John Stewart wrote:

I take from your comments that you would like to run the E182CC at 35 ma
plate current, each side. Max plate dissipation is 4 watts per side.
That results in a max plate voltage of 114 volts.

Looking at the plate family of curves I estimate by eyeball Rp to be
about 2.5K. Calculation at the spec sheet operating point of mu equal to
24 & G being 15 ma/v gives Rp to be 1.6K. But of real concern is the
very short grid base at these conditions while Eb is 114 volts & grid
bias is -ve 2 volts to set that Q point!

In a general sense, the higher the mu of a triode while used as a power
amp then the supply volts needs to be increased in order to develop any
reasonable output power. By eyeball I would think you would need to run
a plate supply of something like 150-160 volts & increase the grid bias
so that Pd of 4 watts is not exceeded.

In any case, it looks like the resulting Rp would be about 3K. In a PP
amp the source then becomes 6K, working into whatever OPT you select.

The 15K/135K interstage transformer I referred to as used by Lloyd
Peppard is a Hammond 124E, the same as you had referred to. Not sure if
you had the Hammond in mind. I have one here that I use now & then in
experimental circuits. But others make something similar, I guess.

When I referred to boutique parts I had in mind your reference to the
HA-106 at $360.00. Fairly obvious to anyone, Hammond is anything but
boutique.

Cheers, John Stewart

In article , wrote:

John Stewart wrote:

I take from your comments that you would like to run the E182CC at 35 ma
plate current, each side. Max plate dissipation is 4 watts per side.
That results in a max plate voltage of 114 volts.

Looking at the plate family of curves I estimate by eyeball Rp to be
about 2.5K. Calculation at the spec sheet operating point of mu equal to
24 & G being 15 ma/v gives Rp to be 1.6K. But of real concern is the
very short grid base at these conditions while Eb is 114 volts & grid
bias is -ve 2 volts to set that Q point!

In a general sense, the higher the mu of a triode while used as a power
amp then the supply volts needs to be increased in order to develop any
reasonable output power. By eyeball I would think you would need to run
a plate supply of something like 150-160 volts & increase the grid bias
so that Pd of 4 watts is not exceeded.

In any case, it looks like the resulting Rp would be about 3K. In a PP
amp the source then becomes 6K, working into whatever OPT you select.

The 15K/135K interstage transformer I referred to as used by Lloyd
Peppard is a Hammond 124E, the same as you had referred to. Not sure if
you had the Hammond in mind. I have one here that I use now & then in
experimental circuits. But others make something similar, I guess.

When I referred to boutique parts I had in mind your reference to the
HA-106 at $360.00. Fairly obvious to anyone, Hammond is anything but
boutique.


Hi John,

I designed the amplifier using the simplified Class A Triode design
guidelines that Patrick posted in rec.audio.tubes a month or two back. I
am going from memory without looking it up, but that gave me a plate
voltage of 162 volts and a total current of 40 mA per channel, 20 mA in
each of the two push pull triode sections used in each channel. I
followed Patrick's recommended plate dissipation of 80% of the rated plate
dissipation, which gave me 3.2 Watts per triode section.

I have decided to initially build only one channel on a chassis sized to
accommodate two channels for stereo. This will allow me to sample the
amplifier with my existing Hammond output transformer, and then if I like
it I can purchase two of the new Hammond 1609 transformers. This assumes
that my existing transformer, once I am able to locate it, is the same
size as the current Hammond 1609 Transformers. If the new 1609 won't
easily fit the same chassis mounting holes, then I will have to rethink
this plan, and either start with a throwaway mono block, which I don't
want to do, or just spring for the two 1609s.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

[John L Stewart]

Originally Posted by John Byrns

In article , wrote:

John Stewart wrote:

I take from your comments that you would like to run the E182CC at 35 ma
plate current, each side. Max plate dissipation is 4 watts per side.
That results in a max plate voltage of 114 volts.

Looking at the plate family of curves I estimate by eyeball Rp to be
about 2.5K. Calculation at the spec sheet operating point of mu equal to
24 & G being 15 ma/v gives Rp to be 1.6K. But of real concern is the
very short grid base at these conditions while Eb is 114 volts & grid
bias is -ve 2 volts to set that Q point!

In a general sense, the higher the mu of a triode while used as a power
amp then the supply volts needs to be increased in order to develop any
reasonable output power. By eyeball I would think you would need to run
a plate supply of something like 150-160 volts & increase the grid bias
so that Pd of 4 watts is not exceeded.

In any case, it looks like the resulting Rp would be about 3K. In a PP
amp the source then becomes 6K, working into whatever OPT you select.

The 15K/135K interstage transformer I referred to as used by Lloyd
Peppard is a Hammond 124E, the same as you had referred to. Not sure if
you had the Hammond in mind. I have one here that I use now & then in
experimental circuits. But others make something similar, I guess.

When I referred to boutique parts I had in mind your reference to the
HA-106 at $360.00. Fairly obvious to anyone, Hammond is anything but
boutique.


Hi John,

I designed the amplifier using the simplified Class A Triode design
guidelines that Patrick posted in rec.audio.tubes a month or two back. I
am going from memory without looking it up, but that gave me a plate
voltage of 162 volts and a total current of 40 mA per channel, 20 mA in
each of the two push pull triode sections used in each channel. I
followed Patrick's recommended plate dissipation of 80% of the rated plate
dissipation, which gave me 3.2 Watts per triode section.

I have decided to initially build only one channel on a chassis sized to
accommodate two channels for stereo. This will allow me to sample the
amplifier with my existing Hammond output transformer, and then if I like
it I can purchase two of the new Hammond 1609 transformers. This assumes
that my existing transformer, once I am able to locate it, is the same
size as the current Hammond 1609 Transformers. If the new 1609 won't
easily fit the same chassis mounting holes, then I will have to rethink
this plan, and either start with a throwaway mono block, which I don't
want to do, or just spring for the two 1609s.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

Hey John Byrnes, did you ever build that Potato Amp?