Yes, you're right that if you can combine a fast amp with a slow one the slow one should be able to use slow transistors. But I'm not sure how well the 405 actually managed to achieve that goal in practice. Never built one, haven't heard one so I can't comment from first hand.
What I can say is that the Bailey 30W sounded a lot better (to me) than many other designs around at the time, and I've always regarded the "inclusive" capacitor as phase lead. Not sure if it really counts as a Miller though.
What I can say is that the Bailey 30W sounded a lot better (to me) than many other designs around at the time, and I've always regarded the "inclusive" capacitor as phase lead. Not sure if it really counts as a Miller though.
From memory JLH considered it to be likely people would own medium impedance pre-amps. Valve designs might be typical. The Leak pre power was connected via a very scruffy piece of cable about two feet long ( 60 cm ). It was 75 ohm. By doing this it worked perfectly, audiophile cable didn't work ( circa 30 ohms ). My brother used to make a nice two transistor NPN PNP buffer ( bootstrapped input, gain if required like JLH set up ) for reordering studios to master 78s using the Varislope. Getting this right is a night and day difference.
John. I have seen Miller frequency compensation done between output transistors and VAS base. If time lag is OK there would be plenty of current and might even provide a little HF assistance. I could imagine very fast transistors ideal. I have built this with very few problems. 60 kHz distortion went from - 54 dB to - 70 dB. It gave stability as one would hope.
Although it might not be successful adjusting the gain can work as a volume control ( oscilloscope required ). A linear pot might be best. You can take it as low as you like if the inverting input. Low Z driver required. I have often thought 250 mV sensitivity as starting point for maximum volume. My whole philosophy is as few active devices between signal sources and speaker. This is a way. I have a little class D amp that works exactly this way. PAM8406.
Hi Nigel - that is what I thought "Miller inclusive compensation" or MIC referred to. Yes, it works well particularly to reduce crossover distortion. It works well over one stage of transistors but if over two (driver+output) it often requires additional compensation on the VAS collector - a couple of caps to ground (often used anyway between driver bases and Vcc or -Vcc). I first saw this proposed by Ed Cherry in his NDFL amp.
Some call that "output inclusive compensation" or OIC which seems more appropriate.
Compensation from the VAS collector to inverting feedback point probably ought to be renamed Input Inclusive compensation (IIC) if we're inventing new names for phase lead.
True, I had put a potentiometer to control the volume, but there was a lot of interference. I left mine without a pot.
There are quite a variety of complete, ready-to-use commercial builds of the JLH'69 design amplifier from China out there and naturally, they all need and have a simple, stereo volume control at the input to make them a stand-alone audio amp. The control is usually a proper looking, ALPS RK27 style part with 2 x 10k audio taper resistances. They seem to work fine with no instabilty either, so what's the nature of the instability and why do so many DIY builds suffer from it 
?
I have a suspicion that the input impedance (i.e. the actual pot. resistance chosen) is too high and the circuitry too simple to compensate adequately over the range of adjustment but I'd like see how others tackle this and any related problems.
?I have a suspicion that the input impedance (i.e. the actual pot. resistance chosen) is too high and the circuitry too simple to compensate adequately over the range of adjustment but I'd like see how others tackle this and any related problems.
The 1996 update of the JLH 1969 circuit included a low-pass filter 4k7 Ohms and 330 pF.
The power bandwidth of the 1969 circuit extended beyond the audible frequency range and did not roll of until somewhere near 200kHz.
The 1969 was my first amplifier project and I moved on to other designs because the treble was response was over the top and relentless.
In the 1996 update the input included RC filtering (4k7 Ohm +330pF) resolved my earlier problem.
The collector-to- base, or emitter-to- base capacitance levels in a small signal transistor are 6pF and 19pF in a BC560. With an input impedance of 2k5 Ohms transistor how much will that do to curtail power bandwidth.
The power bandwidth of the 1969 circuit extended beyond the audible frequency range and did not roll of until somewhere near 200kHz.
The 1969 was my first amplifier project and I moved on to other designs because the treble was response was over the top and relentless.
In the 1996 update the input included RC filtering (4k7 Ohm +330pF) resolved my earlier problem.
The collector-to- base, or emitter-to- base capacitance levels in a small signal transistor are 6pF and 19pF in a BC560. With an input impedance of 2k5 Ohms transistor how much will that do to curtail power bandwidth.
Member
Joined 2009
Paid Member
That agrees with several comments I've read but it hasn't been my experience for some reason. My first build was only about 11 years ago, just soldering a cheap kit together, adding a large heatsink and powering with a bench supply.
It sounded magnificent driving an old pair of Celestion Dittons which were nothing special or particularly sensitive as 1970s consumer grade speakers go. They do have an unusually good combination of sound qualities that suit almost any small audio system, so they still sound quite nice and can't easily be dismissed as cheap trash. I did try more recent and up-market KEF, B&W and local products that I own or have access to but generally, lower sensitivity makes them poor cousins.
I rechecked with a more recent 1969 build where there are a few substitute, fancy parts like 30MHz power transistors, 1µF MKP input cap. and MLCC caps elsewhere but no instability showing on the 'scope nor excessive treble beyond an expected 15 kHz. Maybe I shouldn't describe treble effects at my age but I'm interested in the facts, even though I need a sound level meter, since I can't hear much over 14kHz now.
Next up, I have found and bought a pair of new PCBs for the uncommon 1996, dual supply version. This won't be the first time I set out to compare DC and capacitor coupled JLH versions but unfortunately, the PCBs are for TO3 transistors - not a big problem but my stock semis are all plastic types now so a bit of kitbashing is in order. Here's what the PCBs look like:2PCS JLH1996 Class A Power Amplifier Board 15W PCB|Amplifier| - AliExpress
TIP3055 is an interesting choice. As best I know it's not a 2N3055E or H. The heat sink might need to be slightly larger to help it.
Cable with a 3/4 inch spacing is about 270 ohms RF impedance. Z = root ( L loop/ capacitance open cct )1.4 uH and 49 pF for bell wire or about 170 ohms. The impedance is a ratio so constant. Simple cable like this might sweeten the JLH. Microphone cable is about 33 ohms which might be less good.
Cable with a 3/4 inch spacing is about 270 ohms RF impedance. Z = root ( L loop/ capacitance open cct )1.4 uH and 49 pF for bell wire or about 170 ohms. The impedance is a ratio so constant. Simple cable like this might sweeten the JLH. Microphone cable is about 33 ohms which might be less good.
Low output impedance feedback power amplifier (1966\1963 Ampex Corp)US3246251A
Single-ended push-pull amplifiers (see the SEPP configuration in the "Glossary of terms" section above) and phase-splitters therefor, i.e. the phase splitter circuitry which is suitable for driving them.
US3246251A - Low output impedance feedback power amplifier
- Google Patents
"The phase splitting transistor 22 functions in a conventional manner (see Basic Theory and Application of Transistors, Department of Army Technical Manual, published March 1959, pages and 126) ...The output terminals 27 and 23 of the phase splitting means '20 are connected to a push-pull amplifier means 40 and more particularly to the transistors 42 and 52. The transistor 42 which forms one stage or part of the push-pull amplifier means 40 is connected in a common collector configuration with its base 43 directly connected to terminal 27. The collector 44 of transistor 42 is connected to the E voltage supply or terminal. This E terminal may be suppli d with a 12 volts from any of the well-known conventional sources (not shown). The emitter 45 of the transistor 42 is connected to the amplifier assembly output terminal 65 and'to the feedback point 66 via output current limiting resistor 46.
The transistor 52 which forms the other part of the push-pull amplifier means 40 has its base 53 directly connected to the terminal 28 and is connected in a common emitter configuration. The emitter 54 of this transistor 52 is connected to the +E terminal of supply via output current limiting resistor 55. The +E terminal may be supplied from a conventional +12 volt source. The collector 56 of the transistor 52 is directly connected to the feedback point 66 and the output terminal 65.
It should be understood that the push-pull amplifier means is designed for Class A operation. In accordance with such operation either transistor 42 or 52 will not be made completely nonconduotive during normal operation. These transistors 42 and 52 will alternately be made the dominant output element, that is alternately the transistors 42 and 5-2 will conduct heavily while the other transistor remains relatively nonconductive. This relative conduction and nonconduction is controlled by the phase splitting means 20 which directly supplies input signals 180 out of phase to the base circuits 43 and 53 to alternately bias the transistors 42 and 52 into relatively high conduction states. The utilization of a pushpull amplifier means tends to minimize distortion and allows the transistors to deliver maximum power."
Single-ended push-pull amplifiers (see the SEPP configuration in the "Glossary of terms" section above) and phase-splitters therefor, i.e. the phase splitter circuitry which is suitable for driving them.
US3246251A - Low output impedance feedback power amplifier
- Google Patents
"The phase splitting transistor 22 functions in a conventional manner (see Basic Theory and Application of Transistors, Department of Army Technical Manual, published March 1959, pages and 126) ...The output terminals 27 and 23 of the phase splitting means '20 are connected to a push-pull amplifier means 40 and more particularly to the transistors 42 and 52. The transistor 42 which forms one stage or part of the push-pull amplifier means 40 is connected in a common collector configuration with its base 43 directly connected to terminal 27. The collector 44 of transistor 42 is connected to the E voltage supply or terminal. This E terminal may be suppli d with a 12 volts from any of the well-known conventional sources (not shown). The emitter 45 of the transistor 42 is connected to the amplifier assembly output terminal 65 and'to the feedback point 66 via output current limiting resistor 46.
The transistor 52 which forms the other part of the push-pull amplifier means 40 has its base 53 directly connected to the terminal 28 and is connected in a common emitter configuration. The emitter 54 of this transistor 52 is connected to the +E terminal of supply via output current limiting resistor 55. The +E terminal may be supplied from a conventional +12 volt source. The collector 56 of the transistor 52 is directly connected to the feedback point 66 and the output terminal 65.
It should be understood that the push-pull amplifier means is designed for Class A operation. In accordance with such operation either transistor 42 or 52 will not be made completely nonconduotive during normal operation. These transistors 42 and 52 will alternately be made the dominant output element, that is alternately the transistors 42 and 5-2 will conduct heavily while the other transistor remains relatively nonconductive. This relative conduction and nonconduction is controlled by the phase splitting means 20 which directly supplies input signals 180 out of phase to the base circuits 43 and 53 to alternately bias the transistors 42 and 52 into relatively high conduction states. The utilization of a pushpull amplifier means tends to minimize distortion and allows the transistors to deliver maximum power."
Last edited:
Member
Joined 2009
Paid Member
My father has a pair of Celestion Ditton 22’s and I too was not expecting much from them. However, I was surprised how darn good they sounded!