Note that the the trimpots are positioned as VR1, VR3 and VR2 from the input side to the heatsink side. Make sure you didn't reverse the positions of VR2 and VR3. Measure the resistance across their outer pins to verify.

If the trimpots are ok, then check your R34 and R35 resistors. Measure them to make sure they are still 0.47 ohms (most DMMs are not very accurate with such low resistances, but you should at least be able determine the resistance to within 1 ohm.

If R34 and R35 are ok, then check to make sure you didn't forget to install the wire jumpers at R30 and R31. Also verify that the resistances of R32 and R33 are correct.

[Edited by amb on 12-01-2008 at 07:18 AM.]

It is OK now. The trim pot is at very low resistance. I have measured the one in other ch and it was about 435ohm. Then I calibrated the trim pot to around 400ohm and now it was OK.

The toasted resistor is R32, it appears to be the only thing wrong with that channel? Any thoughts on what might have caused this? The only time that I've seen anything like this was when the outputs of a beta were shorted when the amp was under a very heavy load. Any reason not to simply replace that resistor and run that channel through the initial setup procedure? I've verified that the outputs are not shorted so I'm not worried that there is a persistent fault with the wiring inside the amp.

Thanks in advance for any help.

Nate

[Edited by n_maher on 12-06-2008 at 08:22 PM.]

If this amp doesn't already have it, I recommend installing a zobel network at the output to prevent a possibility of load-induced high frequency oscillation, which could also cause the MOSFET to fail.

Ok, I'll have a go at testing and replacing those parts tomorrow if I have time. Otherwise, there is a zobel on the speaker outputs but not on the headphone output. I'd have to check what the gain is set at but is there an increased potential for oscillation with lower gain configs?

Thanks for the help.

Nate

---

[Quoting n_maher]

there is a zobel on the speaker outputs but not on the headphone output.

---

---

[Quote]

I'd have to check what the gain is set at but is there an increased potential for oscillation with lower gain configs?

---

---

[Quoting amb]

---

[Quoting n_maher]

They should be one and the same -- i.e., both are connected to the amp's output, right? You don't need redundant zobels.

---

No, the speaker outputs are switched off of the headphone jack and the zobel is installed after the switch so it would only be in circuit when the headphones are unplugged. Sounds like it would be a good idea to move the network upstream so it's always in the circuit.

---

[Quote]

Yes, but that is compensated by larger C2-C5 values as you decrease the gain. See the β22 website parts list.

---

Ok, the cap values are as recommended on the parts list so that shouldn't be a problem.

I bought a tube of IRFZ and IRF9Z for matching according to the schematic of your web site. Of course, they can't be matched P to N.

One observation was that the Vgs of the IRFZ is not stable. They could be varied by hundreds of mV when I rechecked them time and again. Is this normal, or did I buy some fake or bad parts?

Thanks.

Now, some International Rectifier parts numbers are available as legitimate second sources from other manufacturers. AFAIK most of those are devices that I-R no longer makes. How well do these compare to the "originals" is an open question.

The Vgs will change with temperature. If you're touching the MOSFETs as you test them, the Vgs will drift, and if you're running significant current through them, the dissipated heat will also change Vgs. You need to be very quick and consistent in your matching technique.

That said, for the β22 you don't have to get too pedantic with matching the MOSFETs. The only devices that should be matched very well are the input JFET quads.

[Edited by amb on 12-11-2008 at 05:20 AM.]

---

[Quoting amb]

I don't know where you got your MOSFETs from, but these are not hard-to-find items, nor are they very expensive or boutiquey, so I doubt that they are likely candidates for fakesters.

---

That's what I thought. So instead of sourcing them from Digikey, I got them locally only.

---

[Quoting amb]

Now, some International Rectifier parts numbers are available as legitimate second sources from other manufacturers. AFAIK most of those are devices that I-R no longer makes. How well do these compare to the "originals" is an open question.

---

That's what I'm afraid of.

---

[Quoting amb]

The Vgs will change with temperature. If you're touching the MOSFETs as you test them, the Vgs will drift, and if you're running significant current through them, the dissipated heat will also change Vgs. You need to be very quick and consistent in your matching technique.

---

I know these. But the IRFZ behaves quite erratically. This time you put it into the socket and it measures 3.02V Vgs. The next few minutes you put the same piece into the rig and it gives you a reading of 3.23V.

---

[Quoting amb]

That said, for the β22 you don't have to get too pedantic with matching the MOSFETs. The only devices that should be matched very well are the input JFET quads.

---

Yes, I know that. But I don't feel comfortable with the behaviour of the IRFZ. The IRF9Z is perfectly OK to me.

My IRF9Z34NPBF, tube reads "Assembled in Mexico"

---

[Quoting Audio Nut]

Question on my 3 channel Beta. When I turn up the volume, the max volume is reached when the pot reaches approx the half way up point. If you go to max, the volume doesn't go any higher than when you have it at the half way point nor does it distort. Is this normal? I have the Alps pot from Amb and it sounds great otherwise. Thanks.

---

"In" comes from your input jacks, "out" goes to the amp board.

---

[Quoting amb]

---

[Quoting Audio Nut]

Question on my 3 channel Beta. When I turn up the volume, the max volume is reached when the pot reaches approx the half way up point. If you go to max, the volume doesn't go any higher than when you have it at the half way point nor does it distort. Is this normal? I have the Alps pot from Amb and it sounds great otherwise. Thanks.

---

Check to make sure you have your volume pot wired correctly. With the pot shaft pointed at you, and the pins downward, it should be (frmo left to right): ground, out, in.

"In" comes from your input jacks, "out" goes to the amp board.

---

Amb, That's the problem. I forgot to mention I originally wired pot backwards, clockwize decreased volume and vise versa. Didn't realize it mattered other than the backwards volume adjust. My bad. Was going to eventually change it but have been too busy enjoying it! Thanks, as always, for the help!

[Edited by audio nut on 12-11-2008 at 07:17 PM.]

So I have decided to build myself a 4-board balanced beta 22. I have some thought about configuration and wiring but not sure if they are good choice. Your help is much appreciated.

1. Headphones Load. I would like to connect the outputs to 2 pairs of XLR in parallel and be able to drive 2 pairs of headphones simultaneously: HD 600 and K 701. I wonder if Zobel network is needed for this kind of load?

2. Gain. Should I use 5x gain per beta 22 board for 10x total gain, or is 2x per board enough? I have a feeling that the K 701 will benefit from a higher gain due to its inefficiency. I mainly listen to modern (i.e. hot mix) records but also listen to classical stuff from time to time. I am going to use 30VDC power supply.

Regards,
ephrank

My source is going to be the emu 0404 USB balanced out. I'm looking at upgrading to Opus or Buffalo DAC next year. I guess I will start with 2x gain and see how loud it can go.

Cheers,
ephrank