G.Kleinschmidt said:
Glen,
you've got to cut down on the coke.
JC posted a zillion times that his amp has 90V rails.
The 50/60MHz duo 2SA1215/2SC2921 would be the prettier choice if it was a bridged design with even more devices.
I don't get the low Vce part either.
The SK devices can do over 15 amps bursts for 10ms at 40Vce or less.
With 90 volts rails, that's 5000 watts peak in a 0.5 ohm load at 40Vce.
jacco vermeulen said:
And the ON Semi devices in question have a SOA indicating 30A bursts for 10mS for Vce ~less than 30V (Tc=25degC). But this is besides the point. The debate is not over peak collector current capability, but the implications of the drop in fT at high Ic - particularly when the output swings towards the rails, producing a low Vce.
Are the Sankens substantially less troubled in this regard at 15A, than the On Semi devices are at only 5A??
jacco vermeulen said:
Well I missed these zillion mentions. But if it has 90V rails then It most likely isn't a bridged design, given it's rated power output.
I take it that it has 9 pairs of devices then (?) and the marketing guy responsible for "18 beta-matched 15 amp, 60 MHz bipolar output transistors" here:
http://www.parasound.com/halo/jc1.php
.....was telling a slight porky.
Regardless of all this, 135A / 9 devices still = 15A peak per device,
so my inquiry still stands.
Cheers,
Glen
G.Kleinschmidt said:
What do you care about marketing fairy tales ?
hFe isn't worth a damn for either of the two devices once beyond 5/6 amps, part of JC's motivation to pick Hitachi MOSFET "Biggens" for the drivers, my guess.
The single key argument JC gave for the Sankens is price wise.
Big advantage of the Japanese devices is their way superior heat transfer from case to sink.
A designer guy i know from overhere tested the Sankens in his designs at elevated die temperature and low Vce values for peak current capability some 20 years ago.
His top model power amp has 14 of the same devices per ch. as used in the JC-1, comparable max output power level but it can be set to a max bias of 2.5 amps.
(msrp of nearly twice as much as the JC-1 though, the JC-1 duo does $13.5K here)
Big problem to compare power devices from the datasheets are the non-uniform test parameters, some things never change.
Chris,
how's about you solder yourself a 10 dollar uP-programmer and open a various source code thread ?
Attachments
jacco vermeulen said:
What on earth does anything you just posted there have to do with the specific issue of fT drop-off at high peak Ic?
You do realise that I'm not actually criticising John's JC-1, don’t you?
We can still build stable high power RET-output amps with high slew rate, bandwidth and peak load current. Right?
Cheers,
Glen
john curl said:
Hi John,
The "roller coaster" ft of the OnSemi devices is not really much worse, if at all, than that of the Sanken devices. Device ft at high currents is a strong function of Vce. Sanken specs their ft at a more comfortable 12V, while OnSemi specs theirs at both 10V and 5V.
As I've alluded to before, I don't think any of these RETs should be operated at 5A in normal playback operation if the output is going to get within 5V or so of the rails, due to the ft fall-off.
However, I think your question is in regard to the protection circuit under short-circuit fault conditions, where one might not trigger the protection circuit until a current of, say 10A was flowing for 1 ms. Under these conditions, I really don't think we care about ft of the output devices, even in regard to global feedback loop stability, since the loop has been broken by the short circuit. Am I missing something here?
The MOSFETs for which I was describing the short circuit protection circuit do not have the ft fall-off problem at similar values of high currents.
Cheers,
Bob
Bob Cordell said:
From the 'Stereophile' measurements of the JC-1:
http://www.stereophile.com/amplificationreviews/774/index6.html
"With continuous drive, the Parasound clipped at 545W into 8 ohms (27.4dBW)—way above the specified 400W. ("Clipping" is defined, as usual, as the power level where the measured THD figure reaches 1%, and is shown in fig.7 as the horizontal magenta line.) With a low-duty-cycle 1kHz toneburst more representative of music, the Halo was a powerhouse. Its clipping power increased by 0.3dB into 8 ohms, reaching 586.5W at 1% THD (27.7dBW, fig.7, black trace), with 1154W available into 4 ohms (27.6dBW, blue), 2255W into 2 ohms (27.5W, green), and no less than 4.2kW into 1 ohm (27.2dBW, magenta). The latter is equivalent to an output current of 64.7A!"
Sounds to me like the peak collector currents of those RET's are getting well beyond 5A each under these testing conditions without much hassle, even when driven into clipping (low Vce).
With a rated slew rate of 130V/uS, it doesn't seem that the JC-1 is tamed with excessive over-compensation either.
Perhaps average Ic is more important than peak Ic?
Cheers,
Glen
G.Kleinschmidt said:
Hi Glen,
Good point. Each device is hitting a peak of about 10A when putting over 4kw on a tone burst into a 1 ohm load. This is a wonderful accomplishment, and I'm not surprised that the devices can physically do it at 1 kHz. But keep in mind, there is no measurement of distortion or slew rate under these conditions. All I'm saying is that this condition is not the condition for which I would limit the normal peak operating current of each RET to 5A or so. The fact that the amplifier is not breaking into oscillation at the peaks where the ft is down just means that John did a good job with his compensation, leaving adequate margin.
Note, however, that into a more realistic load of 2-ohms, still at over 2 kw, John does obey the rule of thumb with regard to maximum current per device, at about 5A peak per device.
Cheers,
Bob
What a bunch of critics! You do 4.2KW, with over 100V/us, with a single sided design, and then we will compare specs.
For the record, the 135A is possible, but I have never tested it there. I don't write the spec sheet or the ads, but I would complain IF the spec were impossible.
'Stereophile tested to 60+ A and that is 4 times what a Halcro could do, if tested the same way. This is a conscious design decision, but not the real reason for the 9 part pairs per channel. The first reason is for increased high voltage safe area, the second reason is for the most effective utilization of the heat sink provided.
I don't get to choose the outside case. I work with what I am given to work with.
Personally, if I were to make my own 100% design, I would most probably use a balanced bridge, similar to what I used as early as 1969, and make an all fet power amp. If this Parasound amp were made in a bridged configuration, it could, in principle, be able to do 1500W into 8 ohms and 2500W into 4 ohms. The biggest problem would be to get power from the wall outlet.
On further consideration of the difference between the ON devices (Toshiba?) or the Sanken, I don't find much difference between the two in change of F(t) with current.
However, what does happen when we reach 10-15A in either device pair? Do we oscillate?
'Stereophile tested to 60+ A and that is 4 times what a Halcro could do, if tested the same way. This is a conscious design decision, but not the real reason for the 9 part pairs per channel. The first reason is for increased high voltage safe area, the second reason is for the most effective utilization of the heat sink provided.
I don't get to choose the outside case. I work with what I am given to work with.
Personally, if I were to make my own 100% design, I would most probably use a balanced bridge, similar to what I used as early as 1969, and make an all fet power amp. If this Parasound amp were made in a bridged configuration, it could, in principle, be able to do 1500W into 8 ohms and 2500W into 4 ohms. The biggest problem would be to get power from the wall outlet.
On further consideration of the difference between the ON devices (Toshiba?) or the Sanken, I don't find much difference between the two in change of F(t) with current.
However, what does happen when we reach 10-15A in either device pair? Do we oscillate?
john curl said:
Who's being critical? Certainly not me, and not Glen, either. If you think either one of us has been critical, please point it out.
Bob
Hi Jacco,
Mostly, I expect I need a mentor to help me with this. Either that or night school.
-Chris
One of the things I have been meaning to get to. I have a 68HC705ICS sitting here that I had planned to teach myself about uP's with. I hooked it up once and the pressures of running a business dragged me away. It might be old news, but I think I should pick up some of those old, but useful uP's and apply them. There are many things an 8 bit uP is overkill for. (I think anyway 😉 )
Mostly, I expect I need a mentor to help me with this. Either that or night school.
-Chris
john curl said:
Thanks for pointing this out. I don't doubt that the 135A is possible, and I salute you for being able to deliver a lot more current and power into low load impedances than the Halcro. Your explanation for the large number of devices makes total sense.
Although Toshiba makes some similarly-numbered devices, I doubt that the OnSemi ThermalTraks are made by anyone but OnSemi. Indeed, I am not certain that the NJL3281D Thermaltrak device is electrically identical in its BJT to the MJL3281A non-ThermalTrak device. I just don't know.
I think too often too much emphasis is placed on the ft of the output devices in regard to global NFB stability and achievable gain crossover frequency. I think that is not really where it is at, as far as why we really want high-ft output devices. The main reason we want fast output transistors is for a very smooth crossover at high frequencies, where the devices are asked to come on and go off while currents are changing very rapidly. This is dynamic crossover distortion, which I believe was first described by Bonjiorno. Another way of looking at it is that we want a virtual total absence of common-mode conduction in the output stage at high frequencies.
In any case, I don't think that the ft of RETs sinking to the single-digit MHz range at 10-15A will cause most amplifier designs to oscillate where the NFB has been designed responsibly.
Cheers,
Bob
ThermalTrak Bias Circuit
This is a ThermalTrak bias circuit that I came up with that I have been working with. It shows a couple of examples of how the ThermalTrak diodes can be used.
The circuit is in the form of a Class-AB 100 watt output stage in simulation form. The circuit is essentially a triple Darlinton “T” circuit that employs two pairs of NJL3281D – NJL1302D OnSemi devices.
Q1-4 just make up a pair of 10 mA current sources that, in combination with the input voltage source, simulate a typical VAS.
Q5-6 and D1-2 make up a feedback-based bias spreader. D1-2 are the key ThermalTrak diodes that work to establish a stable, temperature-tracking bias. The use of the feedback-based bias spreader, as opposed to a conventional Vbe multiplier, eliminates the influence, drift and uncertainty of the pre-driver and driver Vbe’s in establishing the output stage operating point.
Q7-8 and Q9-10 are the pre-drivers and drivers. D5-6 provide a diode drop of voltage offset to make up for the Vbe of Q5-6. D5-6 and Q5-6 are NOT on the heatsink. The relationship of voltage drop for a given current for D5-6 as compared to Q5-6 is involved in setting the bias current.
The voltage drop across R5-6, caused by R7, sets the voltage that will ultimately appear across the emitter resistors R13-16 (albeit modified slightly by some Vbe differences). In practice, R7 would include the bias-setting pot.
The d.c. feedback from the emitters of Q9-10 to the bases of Q5-6 sets the output stage operating point, and forces a tracking relationship between D1-2 and the Vbe’s of the output transistors.
If D1-2 had the same junction drop as Q13-14, and Q5-6 had the same junction drop as D5-6, you can see that the voltage drop across R5-6 would have to equal the voltage drop across R13-14. Corresponding junction drops are at essentially the same temperature. Note that power dissipation, and thus self-heating, in Q5-6 and D5-6 is quite low. This is how the feedback-based bias spreader works. Although in practice the above-mentioned equalities do not hold perfectly, they are close enough, and net differences are made up by trimming R7 to set the idle current.
The two “extra” ThermalTrak diodes, D3-4, are used between the emitters of the driver transistors to establish the idle current of the drivers while keeping the impedance between the emitters very small, so that the drivers can operate in push-pull to provide turn-on and turn-off current to the output transistors. Keeping the impedance between the driver emitters very low at high frequencies is essentially what is often done with the speedup capacitor. The ThermalTrak diodes make it possible to do this in a d.c. fashion with the necessary precision because they track the output transistor Vbe’s.
Cheers,
Bob
This is a ThermalTrak bias circuit that I came up with that I have been working with. It shows a couple of examples of how the ThermalTrak diodes can be used.
The circuit is in the form of a Class-AB 100 watt output stage in simulation form. The circuit is essentially a triple Darlinton “T” circuit that employs two pairs of NJL3281D – NJL1302D OnSemi devices.
Q1-4 just make up a pair of 10 mA current sources that, in combination with the input voltage source, simulate a typical VAS.
Q5-6 and D1-2 make up a feedback-based bias spreader. D1-2 are the key ThermalTrak diodes that work to establish a stable, temperature-tracking bias. The use of the feedback-based bias spreader, as opposed to a conventional Vbe multiplier, eliminates the influence, drift and uncertainty of the pre-driver and driver Vbe’s in establishing the output stage operating point.
Q7-8 and Q9-10 are the pre-drivers and drivers. D5-6 provide a diode drop of voltage offset to make up for the Vbe of Q5-6. D5-6 and Q5-6 are NOT on the heatsink. The relationship of voltage drop for a given current for D5-6 as compared to Q5-6 is involved in setting the bias current.
The voltage drop across R5-6, caused by R7, sets the voltage that will ultimately appear across the emitter resistors R13-16 (albeit modified slightly by some Vbe differences). In practice, R7 would include the bias-setting pot.
The d.c. feedback from the emitters of Q9-10 to the bases of Q5-6 sets the output stage operating point, and forces a tracking relationship between D1-2 and the Vbe’s of the output transistors.
If D1-2 had the same junction drop as Q13-14, and Q5-6 had the same junction drop as D5-6, you can see that the voltage drop across R5-6 would have to equal the voltage drop across R13-14. Corresponding junction drops are at essentially the same temperature. Note that power dissipation, and thus self-heating, in Q5-6 and D5-6 is quite low. This is how the feedback-based bias spreader works. Although in practice the above-mentioned equalities do not hold perfectly, they are close enough, and net differences are made up by trimming R7 to set the idle current.
The two “extra” ThermalTrak diodes, D3-4, are used between the emitters of the driver transistors to establish the idle current of the drivers while keeping the impedance between the emitters very small, so that the drivers can operate in push-pull to provide turn-on and turn-off current to the output transistors. Keeping the impedance between the driver emitters very low at high frequencies is essentially what is often done with the speedup capacitor. The ThermalTrak diodes make it possible to do this in a d.c. fashion with the necessary precision because they track the output transistor Vbe’s.
Cheers,
Bob
Guess you'll have to use your imagination until I figure out how to properly attach a png file. I was suspicious when it didn't show up in the preview.
Some of you will undoubtedly be able to draw the schematic from my description 🙂.
Sorry,
Bob
Some of you will undoubtedly be able to draw the schematic from my description 🙂.
Sorry,
Bob
Hi Bob,
Convert it to a jpg or something (the box has a list of file types). Either that or host the file on a server and link to it. This might be better.
You could also "zip it" and attach the file.
-Chris
Convert it to a jpg or something (the box has a list of file types). Either that or host the file on a server and link to it. This might be better.
You could also "zip it" and attach the file.
-Chris
anatech said:
he can always rename it to whatever.zip and attach here
after download ,everyone interested can rename file to whatever.png