Came across what I think is a clever and very useful idea in this TDPRI.com thread:
https://www.tdpri.com/threads/has-anybo ... nt.863499/My Allen Sweet Spot amp kit came with two test ports for mounting on the rear chassis panel to allow output tube bias measurements without going inside the chassis. I thought that was such a great feature that I installed these ports on a couple of BYOC Brit-series Marshall head clones that I built subsequently. There's one problem, though--to calculate the tube dissipation wattage, you also need to know the output tube plate voltage, and that should really be determined whenever you re-bias, such as for changing output tubes, installing a new rectifier, etc. But to get that value, you have to go back inside the chassis with a multimeter and start poking around with DC voltages running in the 350 - 500V range. I'm always very careful doing this and have (so far, thank you, God!) avoided ever getting "bit"....but it still makes me a bit nervous whenever I do it.
So when I came across the thread linked above, I was very interested. It's a simple enough mod to do, involving installation of a two-resistor voltage divider between chassis ground and the tube socket pin for the plate (pin 3 for all of the common octal types) of one output tube. Then a test port (a.k.a. "tip jack"; see
HERE) is mounted through the chassis and connected to the junction between the two resistors. Here's a simple schematic representation:
Attachment:
External Plate Voltage Measurement Port (Safe!).gif [ 22.56 KiB | Viewed 760 times ]
The selection of the two resistors is critical:
- The resistor connected to the B+/plate voltage (R2) needs to be a high value (1 Mohm recommended) to drop that voltage down to the millivolt level through the voltage divider. This will drastically reduce the danger posed by the high voltage present at the plate. A 3W metal film resistor is recommended for this duty.
- A resistor of 1/1000 the value of R2 is recommended for R1. Not only does this complete the voltage reduction at the junction of the two resistors to a safe level, but it also makes the value of that reading 1/1000 of the B+ value. So what you read on your multimeter in millivolts represents that same numerical value in volts, e.g. a 410mV reading at the R1-R2 junction represents a B+ value of 410 VDC. (I used a 1 Kohm 2W metal film resistor for R1, but you could safely go down to 1/4W, since the voltage at the resistor junction will be well under 1V.)
This second point is why it's recommended that you use 1% tolerance resistors for this application. Even assuming a worst case tolerance match-up on the two resistors, the resulting error would only affect the final plate current calculation by maybe +/- 1mA. However, if you're anal retentive like me, it's easy enough to calculate a correction factor based upon
measured values of R1 and R2. For example, if R1 measure 995 ohms and R2 measure 1.006 Mohms, the correction factor for your meter reading would be:
R1/(R1 + R2) = 995/(995 + 1006000) = 995/1006995 = 0.000988
But since we've converted the meter's mV reading to volts, multiply that value by 1000 to get the correction factor, giving 0.998 in this case. Then
divide your meter result by that factor to give the corrected plate voltage result. So for the 410V example above, the corrected value would be 410/0.988 = 415V.
Hopefully, it goes without saying (though I'll say it anyway) that all this precision falls apart if you use an inaccurate multimeter. So while they may be fine for pedal work, I'm not sure I'd trust one of those $7 Harbor Freight multimeters for amp biasing!
I'll post a photo or two of the installation when I finish it. Right now, I have the two-resistor voltage divider installed, but not the test port; I didn't have a drill bit that could get through the stainless steel chassis of my Allen amp and had to order a couple. But I did validate the measurement accuracy through the voltage divider and got a plate voltage result within 1 volt of a direct measurement at the plate pin.
_________________
“My favorite programming language is SOLDER” - Bob Pease (RIP)
My Website *
My Musical Gear * My DIY Pedals:
Pg.1 -
Pg.2