I recently built an 12AX7/KT88 prototype (I do not have "current" schematics for it, sorry), and I think it sounds good, but it is not very loud. I seem to be about to get around 100dB out of it on-axis from about only a foot away from some crappy desktop speakers. I'm also going for more hi-fi than a guitar amp, so I don't expect maximum gain(and I realize a negative feedback loop is missing, not relevant to the question however), but I'm still trying to figure out a few problems.
I looked up Uncle Doug's tutorial on how to measure power output from it using a 1kHz signal on an 8 ohm dummy load, and from my memory I was only getting a few watts (literally like 3W) out of it.
I'm powering it out of an old iPhone which I'm estimating only puts out about about 250mVrms.
Also my OT is a Hammond 1628SEA from Digikey (https://www.hammfg.com/electronics/transformers/audio/1627-1642.pdf). It's a 5k input impedance to multiple outs(I'm using 8), supposed to be rated for up to 30W.
I understand different frequencies will have different responses, etc, but I wanted to try an exercise in LTspice to see if I could work backward, even if strictly theoretical and possibly unrealistic, and just pump in some raw numbers to see what I would get out.
Here's my slop-matic:
There are many obviously unrealistic things about this schematic, but for the case of modeling, the idea was to swing the grid between -5V and -35V against a 530V B+, using a model(hopefully correct) of the OT matching the Hammond, into a dummy 8ohm load and see what I would get out. In this case, I was getting about 3.2Vrms across R10, which should translate to about 1.28W.
With the OT windings being 25:1 for 8ohm, my guess were that if I wanted to get this to 10x as much (12.8W), the secondary would have to swing 32Vrms, which means I'm at 800Vrms on the primary, which is more than B+ supplied even if I'm going after the opposite swing from the transformer.
I guess what I am trying to do is start a design again, working backward, and figuring out what I need to feed the OT, then what I need to feed the power stage, etc, etc.
I'm sure the things I'm doing wrong are almost innumerable, but if someone can point out a few big hitters, I can get my mind back on the right track.
Thank you!
I looked up Uncle Doug's tutorial on how to measure power output from it using a 1kHz signal on an 8 ohm dummy load, and from my memory I was only getting a few watts (literally like 3W) out of it.
I'm powering it out of an old iPhone which I'm estimating only puts out about about 250mVrms.
Also my OT is a Hammond 1628SEA from Digikey (https://www.hammfg.com/electronics/transformers/audio/1627-1642.pdf). It's a 5k input impedance to multiple outs(I'm using 8), supposed to be rated for up to 30W.
I understand different frequencies will have different responses, etc, but I wanted to try an exercise in LTspice to see if I could work backward, even if strictly theoretical and possibly unrealistic, and just pump in some raw numbers to see what I would get out.
Here's my slop-matic:
There are many obviously unrealistic things about this schematic, but for the case of modeling, the idea was to swing the grid between -5V and -35V against a 530V B+, using a model(hopefully correct) of the OT matching the Hammond, into a dummy 8ohm load and see what I would get out. In this case, I was getting about 3.2Vrms across R10, which should translate to about 1.28W.
With the OT windings being 25:1 for 8ohm, my guess were that if I wanted to get this to 10x as much (12.8W), the secondary would have to swing 32Vrms, which means I'm at 800Vrms on the primary, which is more than B+ supplied even if I'm going after the opposite swing from the transformer.
I guess what I am trying to do is start a design again, working backward, and figuring out what I need to feed the OT, then what I need to feed the power stage, etc, etc.
I'm sure the things I'm doing wrong are almost innumerable, but if someone can point out a few big hitters, I can get my mind back on the right track.
Thank you!
Last edited:
Did you mean the iPhone only puts out 250mV?
(not 250V).
Even a CD player puts out 2Vrms.
That is 18dB louder than 250mVrms - 64 times more power.
You are going to need a driver tube in front of the KT88.
(not 250V).
Even a CD player puts out 2Vrms.
That is 18dB louder than 250mVrms - 64 times more power.
You are going to need a driver tube in front of the KT88.
First idiot mistake. 10 times voltage is 100 times watts, so only an additional ~5.6V swing(140V on the primary). That's the error I was looking for I think.
Thank you.
By law, the phones output is limited for health reason. The phones output is 0.5v peak I think for 0db.
Even with this, I mentioned in the OP that there would be a driver(12AX7 in my first try, and in its place is the voltage source) and the point of the exercise is that, as a voltage amplifier, what would that stage need to output to be able to drive the KT88 to desirable levels. The iPhone alone definitely could not drive that.
That aligns close enough with my practical measurements. I made a major typo in my post. Thank you for correcting me on the power measurements. I have to go back and redo my math to see if it answered my question completely. If you see any other slop in my logic/work please point it out. Thank you again. Clearly I've been looking too closely for too long.
Of course RMS is a tough thing to measure in audio. I just played a Daft Punk song("One More Time", it's what showed up on Pandora in a pinch) on mine on the oscilloscope(hi pot, not connected to the amp) and saw on average around 1Vpp and when the kick hit, around 1.5Vpp. It's an old phone with old settings so it may have an EQ setting on it that I need to readjust. Overall I don't think 250mVrms is that far off, definitely not 2V! I am not sure about a CD player(tabletop vs portable?).
Last edited: