To: “John Vivian” jedvivian@hotmail.com
Subject: Re: TEK P6046

——–

I was able to successfully replace the input pair of FETs with a standard 2N part and get it working well again, except the HF CMRR doesn't seem to be quite up to spec. So these things can be repaired. Are you sure it is the probe as opposed to the amp, or were you using probe to refer to the whole deal? To find a balance problem you should measure DC voltages starting at the input stage and working toward the output until you find the problem.

Oh, and also, check that the supplies are good before you do that.

If this thing has been fiddled with a lot, it may be that an input-side balance adjustment has been twiddled far from it's proper position. So you need to approximately re-balance each stage by the DC level as you work forward. If you can't balance a stage, then that's probably the problem point, though if the stage has a lot of gain you should make sure that it isn't due to too large an input.

I would not assume that the problem lies in a transistor. It could easily be a resistor, or perhaps a bad contact.

The balance on mine is a bit touchy, and I live with a bit of baseline shift between ranges. There is a definite warmup period, and I also find that if, for example, I hold the probe in my cupped hands for 30 sec, there is a detectable baseline shift.

One problem I had is that the calibration procedure calls for using a special version of the probe cover that has more holes, and not having that, I couldn't really do some of the adjustments, as CMRR behavior is significantly affected by the presence of the cover. I thought about drilling more holes in the cover, but didn't.

Rob

To: “John Vivian” jedvivian@hotmail.com
Subject: Re: TEK P6046
Date: Sun, 28 Mar 2004 08:26:26 -0500
From: Rob MacLachlan Rob_MacLachlan@ram2.ius.cs.cmu.edu

I've used hand-matched 2N4416A's to fix both the P4046 and also my 7A13. As I recall, what I did was measure the IDss of the original that wasn't blown. Then I bought a bunch of 2N4416As (20 or so), and tried to find two with similar IDss, then I also looked at the schematic to figure the bias current, and found the pair with the most similar Vgs at that operating current.

You don't really have to match the IDss of the original, but the IDss does has to be greater than the the bias current in the circuit.

In the 7A13, I actually I ended up with a pair whose voltages were better matched than the ones written in the schematic (i.e. the ones in the reference unit originally used to measure the voltages in the schematic.)

Since all the FET parameters are correlated, you'd probably do well if you can just find a pair with well-matched IDss whose IDss is close to the original pair. To measure IDss, just short the gate and source, connect the drain to ~+7V, and measure the current from source to the supply return.

w.r.t. the amp balance adjustment, there's a procedure in the manual for setting it. I think you set it to minimize the baseline shift on the 1-2-5 change, though perhaps not on the lowest decade. Generally for any balance adjustment, there's a particular gain change that that adjustment should be used for.

Rob

cc: “John Vivian” jedvivian@hotmail.com
Subject: Re: TEK P6046

——–

b.t.w., do heed the part of the instructions about grounding the probe before measurement. There is no protection whatsoever on the input FETs. My probe was working just fine when I bought it, but I managed to blow out one of the input FETs in my first hour of playing around with it. I never exactly figured out what I did wrong. Now I use the 10x attenuator when at all possible, and always use the ground clip when I'm not using the attenuator.

To blow the input FETs in the 7A13 is much harder, but I managed to do it by putting it on AC coupling, then repeatedly probing +300V and ground, effectively discharging a 0.1uf cap at 300V into the input.

Rob