It was voltage converted to power at 8 ohms. The peak average power circuit could capture a 10 us peak with reasonable accuracy and would hold it for about 1 second.
Cheers,
Bob
Hi Bob,
In your amplifier designs I have noticed that the VAS bias current is generally set to 10mA. I was wondering what your thoughts were on lowering this from 10mA to 7.5mA or even 5mA once a pre-driver transistor is added.
The reason I thought that it could be of benefit is it will give you more IPS open loop gain if the same 10:1 degeneration of the VAS emitter transistor is used and the current requirements of the output stage should be a lot lower with the addition of the extra pre driver transistor.
I thought that the IPS open loop gain would increase because the input impedance looking into the VAS stage will increase as the VAS emitter resistor is increased.
When you have time I would love to know what your thoughts are on this?
In your amplifier designs I have noticed that the VAS bias current is generally set to 10mA. I was wondering what your thoughts were on lowering this from 10mA to 7.5mA or even 5mA once a pre-driver transistor is added.
The reason I thought that it could be of benefit is it will give you more IPS open loop gain if the same 10:1 degeneration of the VAS emitter transistor is used and the current requirements of the output stage should be a lot lower with the addition of the extra pre driver transistor.
I thought that the IPS open loop gain would increase because the input impedance looking into the VAS stage will increase as the VAS emitter resistor is increased.
When you have time I would love to know what your thoughts are on this?
No. All it did was register squared voltage calibrated to represent equivalent sine wave power into 8 ohms. So, for example, if it detected a 40 V peak, it would register 100 watts.
There is a brief description and block diagram of the peak-average meter on my website at cordellaudio.com under the Audio Instrumentation tab.
Cheers,
Bob
There is nothing really sacred about the VAS bias current being 10 mA, and using an output triple makes it a little easier to justify going lower than 10 mA. However, there is usually an abundance of open loop gain anyway. If you go too low on VAS bias current, you might compromise slew rate, since the VAS must drive all of the capacitances at its output node, including all of the Ccb's and the compensation capacitor. It also must drive the capacitance of the VAS Zobel network if one is used. In my view, the inherent slew rate capability of the VAS itself driving all of this capacitance ideally should be on the order of 50 V/us or more.
Cheers,
Bob
The lateral mosfet drivers are supposed to compensate the termal drift, but I guess they have to be biased more than 25 mA then.
I run 6 pairs J55/K175 on 55 volt rails and 65 volts for the Servo driverboard. Had to fiddle a bit with the OPS bias resistors and ended up with 39R and about 23 mA current, and a total of 900 mA per side for OPS.
Sounds great, about 0,004% distortion in 4 ohms 250w, but I'm chasing a slight hum in speakers with no signal. Will try J200/K1529 for next build.
BR
/Figge
Ah, by instantaneous power I thought you went to the trouble of sensing current too...
Hi,
I think it was this thread, but i cannot locate the post.
I seem to recall that some transistors were better performing than most others for small signal and the VAS of an amplifier, but were in short supply. Can anyone remember the transistors - were they the KSA1381 and KSC3503 (both TO-126)?
I have located the KSA992 and KSC1845 (both TO92) which have Cob=2pF and 1.6pF respectively. The 2N5551 is now not available as TO-92 unless you purchase Farnell's Multicomp Pro transistors. Are there any better transistors to use with a high Vcbo etc. ?
Thanks and regards,
Shadders.
I think it was this thread, but i cannot locate the post.
I seem to recall that some transistors were better performing than most others for small signal and the VAS of an amplifier, but were in short supply. Can anyone remember the transistors - were they the KSA1381 and KSC3503 (both TO-126)?
I have located the KSA992 and KSC1845 (both TO92) which have Cob=2pF and 1.6pF respectively. The 2N5551 is now not available as TO-92 unless you purchase Farnell's Multicomp Pro transistors. Are there any better transistors to use with a high Vcbo etc. ?
Thanks and regards,
Shadders.
Hi rsavas,
Thanks for confirming the transistors for the VAS.
I wanted to move to TO-92 packages as i have had a few issues with regards to the FMBM5551, which is the dual transistor surface mount alternative to 2N5551. When i put the FMBM5551 onto the board, i may be putting too much pressure on the device which causes the device to fail. When i remove the failed device (destructively) i can see the composition where a pressure on a pin could cause internal shifting of the silicon or wire breaking.
Regards,
Shadders.
Thanks for confirming the transistors for the VAS.
I wanted to move to TO-92 packages as i have had a few issues with regards to the FMBM5551, which is the dual transistor surface mount alternative to 2N5551. When i put the FMBM5551 onto the board, i may be putting too much pressure on the device which causes the device to fail. When i remove the failed device (destructively) i can see the composition where a pressure on a pin could cause internal shifting of the silicon or wire breaking.
Regards,
Shadders.
Both KSC1845 and KSA992 are nice devices. I use many of them and I never had any problem with them. The hfe is about 350-400 in my samples, which is much higher comparing with the 120-140 of my Onsemi 2N5551/5401s.
Sajti
That surface mount version is rated at 350mW per transistor compared to 650mW for the TO92 - a limitation of the SOT6 package.
SOT223/SOT89 is my surface mount package of choice for higher dissipation as it can dissipate up to 1.5W with suitable copper-pour in the PCB design. You can also solder it onto a row of 3 0.1" header pins for breadboarding purposes
The DZT5551 and DZT5401 are the SOT223 versions of the 2N5551/2N5401, Digikey stock them.
[ Here's how I add some heatsinking for use with a breadboard:
Last edited:
Hi satji,
The KSA992/KSC1845 do simulate very well, slightly better than the 2N equivalents.
Hi Mark,
That seems to be a good idea for prototyping using surface mount devices. I went for the dual surface mount devices as they are closely matched, but i do need to be careful with the placement so as not to destroy them.
I am going to move to TO-92 packages in some areas to help replacement when required, but the KSA/KSC TO-92 has a different pin out to 2N devices, so the PCB will specific for one or the other (unless i bend the pins).
Regards,
Shadders.
The KSA992/KSC1845 do simulate very well, slightly better than the 2N equivalents.
Hi Mark,
That seems to be a good idea for prototyping using surface mount devices. I went for the dual surface mount devices as they are closely matched, but i do need to be careful with the placement so as not to destroy them.
I am going to move to TO-92 packages in some areas to help replacement when required, but the KSA/KSC TO-92 has a different pin out to 2N devices, so the PCB will specific for one or the other (unless i bend the pins).
Regards,
Shadders.
I am curious, how and why do you need to apply pressure in order to solder this device?
For smt protoyping, I usually specify HASL lead plating.
To mount a leaded device, I put a very small amount of extra solder on one pad (corner pad), position the device over the pads, holding the device with tweezers, re-heat that one pad with extra solder, to tack it in place, then solder all the rest. You have to use very low heat (600F tip) and be fast as not to over heat the device and de-laminate. I use AIM glow-core no-clean eutectic SN63/PB37 solder, 0.015"(0.4mm) diameter.
Some extra info,
The sot-23 version of the device MMBT5551LT3G, has a common footprint
I see that Diodes Inc have the ZXTN5551FL, in a sot-23, it has slightly higher Pd than the above.
The sot-23 version of the ksa992/ksc1845 are FJV992/1845
MMBT5551LT3G data sheet shows Pd of 225mW @ Ta of 25C, on a FR5 pcb
ZXTN5551FL data sheet shows Pd of 330mW @ Ta of 25C, on a 1 sqin high coverage FR5 pcb
Hi rsavas,
Thanks - I can only assume it is the pressure i am placing on the component. I use the same method as yourself, one pad used to secure the device and then opposite corner to ensure the IC is level. I use lead free solder, to ensure ROHS.
For the SOT6, there are two different manufacturers for the 5551, but they have different footprints. I am not too bothered about the power aspect - as i only use them in low power parts of the circuit.
Hi Mark J,
I have a hot air station - but preference is for hand soldering using a soldering iron - it is just as fast to implement.
Regards,
Shadders.
Thanks - I can only assume it is the pressure i am placing on the component. I use the same method as yourself, one pad used to secure the device and then opposite corner to ensure the IC is level. I use lead free solder, to ensure ROHS.
For the SOT6, there are two different manufacturers for the 5551, but they have different footprints. I am not too bothered about the power aspect - as i only use them in low power parts of the circuit.
Hi Mark J,
I have a hot air station - but preference is for hand soldering using a soldering iron - it is just as fast to implement.
Regards,
Shadders.