Tripath-Amp 1-B gain problem

[Bgt]

Have just built this amp but would not switch on because I had some offset(2Volts) so I changed the R33 value from 1k to 900 ohms. Now it works ok but the gain is way of from 1 channel to the other. I don't know if it has something to do with this R33 resistor being changed. The gain of the other channel is 1.5 times lower.
Anyone?

 

[sx881663]

Offset problem

R33 is in the feedback loop and changing it will change the gain but probably not to the extent you are seeing. You need to put it back to what it was and solve the problem by finding the cause. This board has servo circuits to correct offset but they are designed to only cover a small range. What other changes have you made? Did you jumper out the input caps? These caps or off board film replacements are necessary or a lot of offset will be generated and the circuit will indeed shut down. Is the offset in only one channel? If the power supply voltages are very far off that can cause these kinds of problems as well.
On the chip, pins 23 and 26 should both be equal and at 2.5v. If not it has to be a short or a bad chip. The potential cause could be a solder bridge around this area of the chip like 22 shorted to 23 or 25 shorted to 26. This would also change the gain. Use a strong magnifying glass and check the chip pins first then the whole boards for shorts. Use solder wick to remove any and retest.
Roger

 

[Bgt]

Found it!! I placed a wrong value smd resistor on R37, 20k instead of 8k2. Now it works like a charm.
The sound is very nice.
A bit like my UCD180's and 400's.
The only thing...there is a bit of a hiss at a very low volume.
This is the feedback design drawback that the UCD's don't have.
Still...beautiful sound.
For the money it costs....excellent.

 

[Bgt]

the hiss is gone after 5 minutes.

 

[theAnonymous1]

My two completed AMP1-B's also have quite a bit of hiss reminiscent of my SI 5066 but a bit louder. The hiss from mine unfortunately doesn't go away after any amount of time. All in all, it still sounds wonderful.

 

[Bgt]

the longer I listen to them the more I find it unbelievable to have a power amp measuring 23x23x4cm with this excellent sound stage. PSU is a rewound electronic transformer for halogen lamps, 2 double TO220 schottky diodes, 2 caps, BHC 15000uF/40V. Type=150 VA which is enough.
Output is about 58V peak at 8 ohms, both channels driven=52W per channel. Hot? Absolutely not. Only thing that gets hot are the schottky diodes in the PSU which have no heatsinking. But this is only at full power with a 1kc sinewave.
With normal music they stay rather cool.
And the hiss.......strange that it disappears after 5 minutes. Sound is very clean, crisp. Find it hard to believe after being spoiled by the UCD's.

 

[ashok]

Hi Bgt,
You said "..............A bit like my UCD180's and 400's....."
Does that mean that the Amp-1B is lacking in some area like bass or midrange clarity or HF ? Or do you mean that it is only lacking in power ? What are the differences you find ......if there are any.
Thanks,
Cheers.

 

[Bgt]

Hi Bgt,
You said "..............A bit like my UCD180's and 400's....."
Does that mean that the Amp-1B is lacking in some area like bass or midrange clarity or HF ? Or do you mean that it is only lacking in power ? What are the differences you find ......if there are any.
Thanks,
Cheers.

Ashok,
I meant the clarity is enormous, the bass is there, midrange also. On average speakers(Wharfdale Diamond 9.1) it is so difficult to find differences. You're talking about very, very small differences. I have to test them with my better speakers(Driades 91) still.
After listening for hours on end I am truly amazed by this small amp. Maybe the combination of amp. and speakers. I first used an active preamp, Rotel RC 990BX, NAD C162 but am now on the passive preamp again. This is the best combination(for me) because it is the cleanest/most honest emotion wise.
There must be something that makes this amp sound so clean/bright without sssssssss's. Maybe the way the feedback is arranged=before the output filter. Who knows?
The UCD's are also clean but different. Maybe a bit less bright, a bit darker. (and they use AD8620's)
You certainly get very addicted(with all these PWM amps) to music this way.
The only thing I changed were the input caps. Used some Sanyo audio caps, 22uF/16V.
Cy Bert

 

[theAnonymous1]

.......Hot? Absolutely not.......

It is really amazing how cool these little wonders remain. I bought a "1ghz P3" heatsink for my AMP1-B and driving a 4ohm load with 28v rails it barely gets warm, definitely not hot. The 60 ohm resistors I had to add to the output of the SMPS's to stabilize them get warmer than the amps heatsink and they dissipate less than 4 watts of heat.

I'm building a 6 channel amp and have 2 of the 3 AMP1-B's done. Just one more 5 hour session of squinty eyes and solder fumes and I'll be done with the last amp.

 

[Bgt]

Here some pics. of this tiny giant.

[IMGDEAD]http://www.xs4all.nl/~grotel/Photo's/amp1b.JPG[/IMGDEAD]

[IMGDEAD]http://www.xs4all.nl/~grotel/Photo's/amp1b4.JPG[/IMGDEAD]

 

[Bgt]

BTW, the 5 Volts comes from a 78SR105HC 3 legged SMPS IC.
Saves add. wires, mess and efficiency.

 

[theAnonymous1]

Nice job Bgt, I like the the trafo. Its much more space friendly than a toroid.

Where did you get case the lovely little chassis?

 

[Bgt]

Nice job Bgt, I like the the trafo. Its much more space friendly than a toroid.

Where did you get case the lovely little chassis?

Audiokit

Its the GX243

And the weight....light like a heavy feather!!!

 

[Bgt]

ok, redid the layout for better crosstalk/soundstage. Hiss is so minimal that with 92db eff. speakers it is unhearable. Or you have to put your ear on the speaker. If you do this you'll hear a tone and some hiss.

[IMGHTTPDEAD]http://www.grotel.nl/photo/AMP1Bv3.JPG[/IMGHTTPDEAD]

 

[thomaseliot]

Hi Bert,

some question about the sound and comparison with UCD180 (I too have TA2022 in AMP5 and UCD180AD)

In my system UCD sounds higher than AMP5 and is less noisy (that doesn't mean AMP5 is noisy). But overall sound is far better in AMP5 .

So, I suspect that my Amp5 has something less than your AMP1b or your UCD has something less than mine.

With UCD I have the SupplyHG from Hypex (200€!) While you have smd components on AMP1 and better caps, and me trough hole and standard caps. I have Fostex FE206E 96db SPL.

What have you that is not standard from 41Hz? I'm curious about your transformer (how many volts DC?) , the input caps (you wrote 22uF, maybe 0,22uF?), and that cap at the PS input.

Ciao

Thomas

 

[Bgt]

What have you that is not standard from 41Hz? I'm curious about your transformer (how many volts DC?) , the input caps (you wrote 22uF, maybe 0,22uF?), and that cap at the PS input.

Hi Thomas,

Input caps are 22uF.
The orange cap on the psu line =10uF/63V only connected to 1 line+ and to 0V.
The booster cap. on the pcb became 1000uf/16v, panasonic low esr.
+and_ line on the pcb became 680uF/35V ultra low esr types, Rubycon.
psu lines are 31V DC each at idling. Halogen tranf. is 150VA/40khz.

Sound is impressive for this tiny amp. No bass or treble missing here.

 

[thomaseliot]

Hi Thomas,

Sound is impressive for this tiny amp. No bass or treble missing here.

Hi Bert,

I know . You have 30.000 uF in the PS, too. I'm interested to your transformer: is it standard? I only found +12V halogen trasf. Is it stabilized?

Ciao

Thomas

 

[Bgt]

Hi Bert,

I know . You have 30.000 uF in the PS, too. I'm interested to your transformer: is it standard? I only found +12V halogen trasf. Is it stabilized?

Ciao

Thomas

It is a rewound transformer inside. Will try to find the pic.
It is not stabilized. It has a 50Hz output and the 50hz cycle is made from a 40khz sq. wave.
[IMGDEAD]http://www.grotel.nl/photo/electronictransformer.JPG[/IMGDEAD]

[IMGDEAD]http://www.grotel.nl/photo/electrans.JPG[/IMGDEAD]

 

[thomaseliot]

It is a rewound transformer inside. Will try to find the pic.
It is not stabilized. It has a 50Hz output and the 50hz cycle is made from a 40khz sq. wave.
[IMGDEAD]http://www.grotel.nl/photo/electronictransformer.JPG[/IMGDEAD]

[IMGDEAD]http://www.grotel.nl/photo/electrans.JPG[/IMGDEAD]

Hi Bert,

It is a switching circuit! And it costs 13 €! You made a smps. I've heard that Tripath chips sound better with smps.

Is it something doable what you made to original Conrad electronic transformer?

I think many people would be interested

Ciao

Thomas

 

[rt7085]

Tone

Somewhere you said that there was a tone in the noise. I bet ya that is because the switching frequency of the supply is not synchronized to the switching frequency of the amplifier. Perhaps you should somehow sync it up, if possible. It is most likely a beat frequency.