[AudioCats]

test result of using a separated power supply for filament (the on-bard supply only does HV/B+)

I was hoping to solve the hum problem by using a dedicated filament supply, but it did not work. There was audio frequency noise in the HV line, some kind of oscillation maybe, (not just the 63khz sharp spikes). It seems that the onboard power IC is expecting the 12V line to have some kind of load, when the current draw goes up the hum/noise frequency increases, and wise-versa. The oscillation was around 10hz when the load resistor was 68 ohms (around 180mA draw). At 0mA draw it became some kind of motor-boating noise.
The "sweet spot" (relatively speaking) happened when the load resistor was 12 ohms (1A draw), but the hum was still noticeable.

 

[baronbeehive]

Not heard of this before, sounds like you've got a faulty unit. Could be a faulty SMPS.

I've had mine for years with no problems.

Edit: Have you tried changing the tubes, one might be faulty?

 

[AudioCats]

sorry, I probably need to clarify: the problem happened when using a separate power supply for the 12v filament power.
The amp works fine in original configuration (but has hum when using sensitive headphones).

 

[baronbeehive]

sorry, I probably need to clarify: the problem happened when using a separate power supply for the 12v filament power.
The amp works fine in original configuration (but has hum when using sensitive headphones).

OK, great! I can't comment as mine is the speaker version, don't know what headphones others have here.

 

[AudioCats]

I recently bought a Mini2013 (four tubes, two 6j1 pushing two 6p1). Sounds pretty ok pushing orthodynamic phones.
Info / notes in the DIY section: https://www.head-fi.org/threads/appj-mini2013-tube-amp-6j1-pushing-6p1-3w-per-channel.966704/

 

[AudioCats]

looks like the mini2013 is another one of appj's dead end design, easy to fail under certain (though unusual) scenarios. The HV rail can drift up to 360v when only 6p1 is plugged in, this voltage can fry both the tube and HV power supply caps.

Thus it can not be recommended due to this potential reliability issue.

 

[AudioCats]

pushing a pair of HE6se, the modded 1502A sounds quite a bit fuller, more ambient and more spacious than the modded Mini2013.

Both have FT aluminum/teflon interstage caps. Input tubes: 1502A has amprex 12ax7, mini2013 has RCA 6ak5.

(PA1502A now uses a transformer in the wood box below to provide both HV and filament power)

 

[baronbeehive]

Just noticed more activity on this thread after a long while!

@AudioCats I hadn't noticed the new(er) Mini2013 before. That voltage sounds a tad high as you say. And with the SMPS which can fail catastrophically there could be issues. But I have to say I've had no reliability issues so I can only assume that the APPJ's are OK from that standpoint, despite a dubious build quality. Have you bypassed the SMPS with your transformer mod on your 1502A?

Re: my APPJ I've had no need to look further as I still have it and the mods have taken it to top level, it's at the same level as my heavily modded Little Dot MKVi+ so no need to change now. To get to top level it needs top quality parts such as Jupiter caps but even so it's still value for money.

I'm interested that you've tried film coupling caps. When I changed out my lytic cathode bypass caps on my LD for film caps the sound flowed brilliantly and the bass really tightened dramatically. At the same time I changed out some coupling caps to 1uF which filled out the sound very much more and increased dynamics also.

For the APPJ I now use a 12AT7 Brimar Footscray factory driver tube and the sound has become more powerful. Re: cathode caps you could go up to1000uF without issue.

I'm quite impressed that your APPJ can drive your HE6se, must say despite only a 3.5w output mine has no trouble driving speakers.

I notice you have done the diode mod. What is this? I've tried to find out without any luck. What effect does this have?

I think it's a shame that APPJ in their infinite wisdom decided to supersede the 1502A and PA0901A range.

 

[AudioCats]

I am not aware of the "diode mod". The only place I can see to put in a diode of some kind, in the signal stream, is to replace the cathode resistor of the input tube with a LED. In the 1502a that resistor drops about 2v which is within the range of LED forward voltage drops. Use different color leds to get different voltage.
But if that is the case, why would they not call it the "LED mod"? or maybe they want to use a string of germanium diode (1n34 etc) to obtain the same voltage drop?

The inexpensive mini2013 has a curious overall arrangement. The tubes used have lower max voltage rating (6ak5 180v max, 6p1p 250v max), while the HV rail voltage is high at 260v. During (every, normal) power-up process, the HV rail drifts to 360v and stay there for a couple of seconds, (filaments not yet hot so tubes not conducting to sag the HV rail), yet for some reason appj used 350v rated capacitors as HV reservoirs. Why didn't they just use the same 400v cap as in the 1502A......

 

[AudioCats]

almost forgot..... output transformer upgrade attempt:

I bought a pair of "Luge amorphous output transformer" and tried it with the 1502A. The transformer is 3.5k/8R so the turn ratio is about the same as the stock transformer which is said to be 20:1.

-- the core is legit metglass. Ultra-thin with that glassy shine.
-- these transformers can not be used as-is, they must be potted or they will "sing" like little speakers. (you hear what is being played even if the phones are not plugged in....).
-- the sound, tested with T50rp (these was before I bought the He6se): maybe a little sharper and more transparent. The difference was not immediately obvious.

So it looks like the little output transformers in the appj are pretty ok to begin with and there is no obvious upgrade path for the output transformers without spending a lot of $.

 

[baronbeehive]

I am not aware of the "diode mod". The only place I can see to put in a diode of some kind, in the signal stream, is to replace the cathode resistor of the input tube with a LED. In the 1502a that resistor drops about 2v which is within the range of LED forward voltage drops. Use different color leds to get different voltage.
But if that is the case, why would they not call it the "LED mod"? or maybe they want to use a string of germanium diode (1n34 etc) to obtain the same voltage drop?

The inexpensive mini2013 has a curious overall arrangement. The tubes used have lower max voltage rating (6ak5 180v max, 6p1p 250v max), while the HV rail voltage is high at 260v. During (every, normal) power-up process, the HV rail drifts to 360v and stay there for a couple of seconds, (filaments not yet hot so tubes not conducting to sag the HV rail), yet for some reason appj used 350v rated capacitors as HV reservoirs. Why didn't they just use the same 400v cap as in the 1502A......

Oh right, I saw your "signal-diode mod comment and assumed that was it. Very early on there were comments about this diode mod but since then I couldn't find out any information about it. I assume it was simply replacing with a faster acting one, as the amp is hybrid and already quite fast I didn't think it was necessary anyway, certainly no mention of LED's.

I don't know but must assume that the specs of the caps used would accommodate higher voltage spikes on start up otherwise they're just asking for trouble. Pretty weird!

So it looks like the little output transformers in the appj are pretty ok to begin with and there is no obvious upgrade path for the output transformers without spending a lot of $.

No, unless it was redesigned but why bother. I believe the OT's are one reason the amp is so good, that and the hybrid design. When I bought mine I knew from the reviews that it was good but didn't realize it could be pushed to the limits that it has, so heads up to APPJ for their little miracle!

 

[AudioCats]

...
... I believe the OT's are one reason the amp is so good, .....

yeah, those are some interesting little transformers.
From what I can see in the mini2013, the output transformer core appears to be just normal stacked EI without a deliberate air gap. Maybe they are counting on the air gap, which is naturally there between E and I pieces, to be close enough.

 

[baronbeehive]

@AudioCats don't know if you are thinking of any further mods? I think it might be interesting to change the driver tube operating point to see if it can be optimised, also see what effect having less GNFB in the feedback loop would have. Someone earlier in the thread said he was going to try to change the GNFB, don't know if he succeeded. As the amp already has good OT's to control the output and as you have put in film caps I think there would be some scope to try the GNFB change without issue. I'm not doing anything further myself, already very happy with my amp.

 

[AudioCats]

after adding the wood box under the 1502A, it is no longer easy to work on the circuit board, at least the underside of it... I think I am done modding on it.

I just noticed that the 1502A is currently available on aliexpress for around $200 (without tubes, which usually gets replaced by NOS anyway). Not that I would recommend any of the appj amps as in my opinion they all need a bit of work and there is not enough room in that tiny case to do much.

If one doesn't mind working on the amp, maybe this clean looking 12ax7/6v6 amp can be a good starting point, don't know how it sounds but the transformers look nice and the top plate layout looks pretty presentable. The wood box is a bit big ( 12"x8") but that also means there will be space to add upgrade parts.
Or a JBH el-34 single-ended for about the same amount of $.

 

[rdale]

Recently I've been using my PA 1502A to drive Grado SR80x's and that was a really fun combination. But I just got some new earpads for my Audeze Sines from vesperaudio. With the APPJ I think this is my favourite combination so far. I'm using an iFi iDSD Micro Black Label as a DAC instead of driving the Audezes directly as I'd only tried before. The pairing is magical and I'm finding it difficult not to listen to 'just one more track':

The amp is sitting on an SSC stand, and I've upgraded the feet, but everything else is stock.