I'd love a list of errata that may be encountered when working with particular devices, and circuits with a particular function. I know that any list, as described, would be built from a solid foundation of electronics theory. With a tube amp powered up, scope testing produced interesting results. A voltmeter probe was touching a preamp tube plate lead, and the scope output trace got "fizzy". The amp has problems, but appears to have correct voltages everywhere. Fizzy, meaning the sine wave turned into a positive DC biased, relatively HF signal. In another instance, on a different part of the circuit, the CLEAN trace became a noisy, fat trace when the voltmeter was added to the circuit. What's up with this?
The long meter lead is an unintended antenna. You are picking-up radiation from other stages and injecting them into that plate and the next stage. Output stage plates have huge swings. When antennaed back into earlier stages, the amp can oscillate. Police/taxi radio signals also. And all our digital electronic toys. CFL and LED lamps.
I would not say "cheap leads". In the old days a VTVM was calibrated for 1Meg *in the probe*, so you were only adding an inch of antenna, not 3 feet of wave-catcher. This idea seems to be forgotten.
Re-route your meter lead away from obvious racket-makers. Some experimentation needed.
I would not say "cheap leads". In the old days a VTVM was calibrated for 1Meg *in the probe*, so you were only adding an inch of antenna, not 3 feet of wave-catcher. This idea seems to be forgotten.
Re-route your meter lead away from obvious racket-makers. Some experimentation needed.
Circuits don't have "errata"; these are errors in printing or writing.
There is a general rule in electronics: change the circuit and you change the circuit's behaviour. Adding any form of test equipment will change the circuit. You need to ensure that you change the circuit as little as possible, in ways which you understand as much as possible. At the very least, test probes add capacitance. They may add coupling to other circuit nodes too. Test probes can make a circuit oscillate, or they can stop oscillation.
There is a general rule in electronics: change the circuit and you change the circuit's behaviour. Adding any form of test equipment will change the circuit. You need to ensure that you change the circuit as little as possible, in ways which you understand as much as possible. At the very least, test probes add capacitance. They may add coupling to other circuit nodes too. Test probes can make a circuit oscillate, or they can stop oscillation.
The term "errata" should really be "phenomenon". As you say DF96, it's not a mistake that these things are happening. I think the problem is with my test setup? It happened again on a DIFFERENT amp. Here is a recap of the issue: Amp is running, current draw is normal, NO INPUT SIGNAL. At this point, I am looking for voltages, plate/grid/cathode. When a standard 10 MEG meter is placed between chassis and the first stage plate lead, a large signal appears on the top half of the scope screen, and the current needle on the variac jumps up quite a bit. The scope signal is taken across the dummy load. The only things on the bench, and turned on are the scope, amp, a Heathkit variac/isolation transformer and two led "bulb base" lamps up above. Testing with a second voltmeter produced the same results.
Yes, I have seen oscillations when certain scope probes or DMM leads are connected, etc.
I agree, a 10X probe with 10 Meg DC loading, and about 2-15 pF (AC/RF loading), are far superior to using a 1X probe (100pf).
And, be careful when using AC coupling on the scope. That is typically good for 300VDC, and beyond that the scope AC coupling cap is leaky or will even be destroyed.
A P5100 100X probe has a series resistor, and a shunt resistor, so that the DC reaching the scope does not exceed 300V for the rated 2500V at the probe tip. Of course there is a lot more complexity of the probe internal circuitry, so that it is good all the way to 75 MHz.
I hope you are not using the isolation transformer to float the amplifier. And I hope you are not using it to float the scope, or any other piece of test equipment.
Having worked on very special floating test equipment setups for years, and knowing what has been designed by a very good knowledgable engineer, and what has not, I know most of these setups cause more trouble than they solve.
Not to mention that I advised companies not to do floating measurements, since the death of a person would mean that the surviving spouse now owned that company.
If you are using the heathkit to isolate the amp, that is also likely part of the problem.
Most isolation transformers have a fair amount of capacitive reactance across the transformer windings, and if not at 60Hz, at least the ground shield (if there even is one) between primary and secondary is Not ground (not at high frequencies and RF, where oscillations occur).
Grounds are Commonly Misunderstood (**).
I had a customer that complained about a 100V signal only 1 foot up the "Ground" return wire. A fast rise high current spike was responsible (ground wires should be measured in NanoHenries per inch).
Better yet rate ground leads in Volts per (amp/nanosecond rise time).
That might remind people of the above (**)
I leave all my Tuners and CD players use the 2 wire power cords they came with.
I Modify all my Power Amps to use a 3 wire IEC connector and 3 wire IEC power cord.
The ground loops between the Tuners or CD players to the amplifier are non existent or of no significant effect, as long as I use good quality phono plug shielded cables for the signals to the amp.
I agree, a 10X probe with 10 Meg DC loading, and about 2-15 pF (AC/RF loading), are far superior to using a 1X probe (100pf).
And, be careful when using AC coupling on the scope. That is typically good for 300VDC, and beyond that the scope AC coupling cap is leaky or will even be destroyed.
A P5100 100X probe has a series resistor, and a shunt resistor, so that the DC reaching the scope does not exceed 300V for the rated 2500V at the probe tip. Of course there is a lot more complexity of the probe internal circuitry, so that it is good all the way to 75 MHz.
I hope you are not using the isolation transformer to float the amplifier. And I hope you are not using it to float the scope, or any other piece of test equipment.
Having worked on very special floating test equipment setups for years, and knowing what has been designed by a very good knowledgable engineer, and what has not, I know most of these setups cause more trouble than they solve.
Not to mention that I advised companies not to do floating measurements, since the death of a person would mean that the surviving spouse now owned that company.
If you are using the heathkit to isolate the amp, that is also likely part of the problem.
Most isolation transformers have a fair amount of capacitive reactance across the transformer windings, and if not at 60Hz, at least the ground shield (if there even is one) between primary and secondary is Not ground (not at high frequencies and RF, where oscillations occur).
Grounds are Commonly Misunderstood (**).
I had a customer that complained about a 100V signal only 1 foot up the "Ground" return wire. A fast rise high current spike was responsible (ground wires should be measured in NanoHenries per inch).
Better yet rate ground leads in Volts per (amp/nanosecond rise time).
That might remind people of the above (**)
I leave all my Tuners and CD players use the 2 wire power cords they came with.
I Modify all my Power Amps to use a 3 wire IEC connector and 3 wire IEC power cord.
The ground loops between the Tuners or CD players to the amplifier are non existent or of no significant effect, as long as I use good quality phono plug shielded cables for the signals to the amp.
It is also possible that your amp is unstable and on the verge of oscillation, adding the meter probe capacity is enough to disturb the circuit and start the oscillation, which can be a very high (UHF or higher) frequency your scope can't display due to its limited bandwidth. The current jump you noticed on your Variac is also a (possible) visible indication of this behaviour.
This is not unusual and well know by (older) techs. For this same reason you can find the following notice on some Audio Research schematics:
"WHERE NOTED, MEASURE VOLTAGES THRU 1K RESISTOR AT TEST PROBE TIP, TO AVOID HF INSTABILITY"
This is not unusual and well know by (older) techs. For this same reason you can find the following notice on some Audio Research schematics:
"WHERE NOTED, MEASURE VOLTAGES THRU 1K RESISTOR AT TEST PROBE TIP, TO AVOID HF INSTABILITY"
Last edited:
Two possible issues:
1. local parasitic oscillation in the stage being measured
2. loop instability triggered by extra phase shift from the probe capacitance
Both can be eased by putting a resistor at the tip. I would use 1M, and then remember that all measured voltages need to be multiplied by 1.1. Or 100k and then you don't need to multiply unless you are really fussy.
1. local parasitic oscillation in the stage being measured
2. loop instability triggered by extra phase shift from the probe capacitance
Both can be eased by putting a resistor at the tip. I would use 1M, and then remember that all measured voltages need to be multiplied by 1.1. Or 100k and then you don't need to multiply unless you are really fussy.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.