You could be right.At the anode of the 6C4 is something like 3Vrms needed for max.output.The cathode resistance over there isn't decoupled,not much amplif.but still allmost 10x.So less than 1/2V sensitivity at the vol.pot.No problem for most sources.
A list of anode+cathode voltages would help to pinpoint a possible fault.
Mona
Modern with some errors
The cathode resistor of the concertina not 22k but 2k2,point missing
Also not the NFB at the cathode but at the 220 ohm ,where the PFB is feed in.With 3k9 // 1k8 you don't get the -1V on the grid.
Mona
Yes, that is a mistake. It couldn't be 22K for that current.Modern with some errors
The cathode resistor of the concertina not 22k but 2k2,point missing
You have to take into account the total current through the 220R resistor as well....
Have another look, it's basically a perversion of a mu follower with differential outputs. I'm pretty confident it works, and it is actually rather clever IMO.
If your's, and Ketje's, idea came true, both outputs only will be symmetrical if V2B's impedance would be infinite. How confidential is this?
I think that anyone capable of PSpice shoud do a simulation to get close to reality.
Best regards!
@45,
Possible the current fom the cathodyne gives the missing extra voltage,didn't check.
It's also the bad feeling of having that chemcal in the current path of the NFB.I don't like that.NFB a the 220 ohm and,if nessesairy adjusting the cathode resistor to get Vg=-1V for the right Ia,gives a much smaler current in that C.
Mona
Possible the current fom the cathodyne gives the missing extra voltage,didn't check.
It's also the bad feeling of having that chemcal in the current path of the NFB.I don't like that.NFB a the 220 ohm and,if nessesairy adjusting the cathode resistor to get Vg=-1V for the right Ia,gives a much smaler current in that C.
Mona
I'm working on a similar circuit which does work perhaps not as well as I would like, and might get around to simulating this one. I suspect the AC balance is not the best in this particular implementation.
The biggest problem with this design and its variants is the relative modest level of open loop gain.
Well no,we have a rather big open loop gain,thanks to the PFB.
Let's round-off for ease.Gain µ-follower =µ12AX7=100.So 1V in gives 100V out.That's feed back with a divider 1k = 150k ,feedback voltage 100/150=.67V .From the outside we need 1-.67=.33V to get 100V out.Gain is 100/.33= 300 ! Maybee only 200x in reality,not that smal
.Mona
Spice Analysis
The performance of this design is surprisingly good based on the Spice analysis I did in LTSpiceIV. Note that I have omitted the compensation components in the feedback loop because I modeled with a near ideal transformer since I know nothing at all about the original. I did however add a couple of hundred ohms of DCR to the inductor models used to make the OPT. Note the 1 ohm resistors in the screen circuits are so that I can measure current.
Output power level used here is ~ 1.0W into 8 ohms, and with a good OPT excellent distortion numbers might be attained. I am working on a similar circuit and will post here later if there is interest. Something on the order of 10 - 12W should be attainable.
The drive phases are not perfectly matched but not bad with feedback applied. Note that if the bootstrap works as well as intended there is almost no audio current through R1 the plate load resistor for the voltage amplifier stage and it does not badly unbalance the concertina phase splitter. (There is obviously an effect on the DC conditions)
I did have some issues with the output stage, something about 75 ohms DCR in parallel with one of the primary of the OPT... Oops..
The performance of this design is surprisingly good based on the Spice analysis I did in LTSpiceIV. Note that I have omitted the compensation components in the feedback loop because I modeled with a near ideal transformer since I know nothing at all about the original. I did however add a couple of hundred ohms of DCR to the inductor models used to make the OPT. Note the 1 ohm resistors in the screen circuits are so that I can measure current.
Output power level used here is ~ 1.0W into 8 ohms, and with a good OPT excellent distortion numbers might be attained. I am working on a similar circuit and will post here later if there is interest. Something on the order of 10 - 12W should be attainable.
The drive phases are not perfectly matched but not bad with feedback applied. Note that if the bootstrap works as well as intended there is almost no audio current through R1 the plate load resistor for the voltage amplifier stage and it does not badly unbalance the concertina phase splitter. (There is obviously an effect on the DC conditions)
I did have some issues with the output stage, something about 75 ohms DCR in parallel with one of the primary of the OPT... Oops..
Attachments
According to Spice transient analysis open loop gain is about 270 or 48dB, clearly OLG is not sufficiently high that you can totally ignore it in gain calculations, and as expected CLG is about 0.65dB lower than you would expect based on the feedback components. Output loaded.. (Note I have not bothered to do an AC analysis as of this point, just lazy and since I don't have a model of the real OPT not that informative either.)
You were not too far off the mark Mona!
You were not too far off the mark Mona!
This schematic, Ketje, shows an arrangement that meets my previous suspicion exactly: The voltage amplifier's plate resistor R5 is connected to plate supply instead of the node below the cathode, as in the OP's original schematic.
Best regards!
I will probably do an analysis of the Philips design in the next few days, what I can say is the combination of bootstrapping and 1% or so of positive feedback results in some fairly impressive open loop gain numbers, balance is pretty good, and can be trimmed if desired, and distortion is very low. I'm working on a design based on these concepts (basically I am ripping this design off.. lol) and despite being an SE guy I think I might just have to build it, I'm pretty sure the performance will be good.
For that, sir, I just listened to 12/31/95 on this setup. Which reminds me again why a single channel is still useful -- there is plenty of music worth listening to with a single speaker.
This weekend was a bust for working on this particular amp due to others that needed more immediate work, but I will get to it soon enough. My next steps are to take the LP/RIAA/78 switch out of the chain, replace the ceramic disc caps where necessary, and search out obvious places for hum as have been suggested here. Once I've had time to do that, I'll start thinking about mods that will make it more functional over all.
In the meantime, I think I'm going to order a single Betsy from Wild Burro. Looks like an OB loaded with a full range driver could be interesting, and this is an easy way into that experiment.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.