Neil Young's "Magic" 5E3

[Helmholtz]

Curiously enough when I measure 5E3 OT I have (Mercury FTDO-59), I get the exact same turn ratio as the 18W (n=34:1 @ 8 ohm).

What test frequency are you using?

Did you try the heater voltage method?

[Mike K]

What test frequency are you using?

Did you try the heater voltage method?

I used 1kHz.

I actually tried multiple frequencies on this one, and I was getting a slightly higher impedance at low frequencies.

I have not tried the heaters yet.

[Helmholtz]

I used 1kHz.

Voltage transfer ratio should not change with frequency within reasonable limits.

What's the bandwidth of your DMM?
I found that both bandwidth and accuracy drop at mV levels with most DMMs.

Most DMMs are quite accurate at line frequency and the higher voltages with the heater method avoid low level errors.

Wondering about your high Raa results with 3 different OTs.
I just measured a Fender Princeton/Super Champ OT (022913). Voltage ratio is 31.6 corresponding to an 8k primary without DCR.

[Mike K]

I think it may be the precision of the DMM. I just measured another OT (SE this time) and I get higher than spec. It's supposed to be 12K, I get more like 13.5-14K, depending on secondary. I get consistent results but I don't have a dedicated mV AC readout, only for DC.

When I measure the 5E3 OT with just one half of the primary, I get more like 8.9k (and multiply by 4) vs. 9.2k for the whole primary.


Typically when I've measured before I've used my variac and dialed it up until I had 1 VAC on the secondary, then I'd measure the voltage on the primary side.


I also measured a couple power transformers using this method, but they have a much smaller turn ratio. And I knew what they were, and I got the right result, so it's likely just the limits of the mV range.

[Mike K]

OK - with heater on the primary, I get n = 32.4 for the 5E3. That equates to an 8.4K primary.

[Helmholtz]

OK - with heater on the primary, I get n = 32.4 for the 5E3. That equates to an 8.4K primary.

Shows that your results significantly depend on (secondary) voltage level.

My actual suggestion was feeding to the secondary.

[Mike K]

My actual suggestion was feeding to the secondary.

I know - but this was much quicker and safer than having to cap everything off and be wary of hundreds of volts on the primary. I concur the results would be much more precise if I were to use voltages more in the range of my meter. I was skeptical of this because with only 1.5V on the primary, you're less than 100mV on the secondary. My meter reads out to 3 decimal places on the AC, but I have no idea what the actual precision is.

[Helmholtz]

My meter reads out to 3 decimal places on the AC, but I have no idea what the actual precision is.

Number of digits or resolution has nothing to do with accuracy.