I could use the schematics for a SoundStream TA1.3000D Tarantula. Also some advice. The customer said that it would play and then just shut down. I have played it for several minutes without any problems at low to mid power. I am going to do some high power testing next. Any fore-knowledge of what to look for would be appreciated.
To all you guys and girls on this thread. I have been away for sme time due to my wife's health issues but thank goodness she is back at 100%.
Secondly to ALL members on this forum, I value all of you, newbies to the most technical. You all make it worthwhile.
Please understand that this reply is not bashing anyone, just my perspective as a design engineer and manufacturer of amplifiers.
Wade it is good to meet you at last. I tried to purchase Soundstream when they were on their way out but my bid was not accepted.
I can only write what I have experienced with SS amplifiers. I look at amplifiers through an engineer's lense.
The orginal D series in my opinion were not very good amplifiers. A self oscillating supply using TIP35 devices is/was not my first choice of power supply design. There is ZERO dead time in these supplies and running at 20KHz is also not optimum. The amplifiers used TO218 darlingtons [TIP142/147] pairs with 0.1 ohm emitter resistors. These and all other plastic darlingtons have big thermal issues and if biased a little too hard, they will run away thermally.
Reference series were a step forward using a SG3525 as the SMPS control chip but the supplies still had issues. I have experienced transformer failures as Perry pointed out. The largest parts failure I have experienced is the rectifier diodes [FEP16/FEN16] where they are paralleled. It is a golden rule not to parallel silicon diodes withour current sharing resistors in the AC legs. Solution use TO247 diodes and use just one pair! Of course the darlington output devices still fail.
Did I mention that to repair these Reference amplifiers one must remove a zillion screws, carefully remove the PCB from the chassis (Oh those with the muliple LEDs to put the PCB back into the chassis), fix the fault and go through the same gymnastics to put the PCB back into the chassis. A time consuming process!
The amplifiers all suffered the same problems that the darlingtons are well know for. To SSs credit they did eventually change to 1 ohm + diode.Without a high value emitter resistor [1 ohm] and a high speed bypass diode, the devices WILL fail if the amplifier's bias spreader voltage is anywhere close to the 2.25-2.35v. NOTE: I always set idle current using three methods.
1] I keep a volt meter across the bias spreader.
2] I drive the amplifier at 20KHz at about 0.5v RMS into a 2 ohm load
3] I look at the THD on my Audio Precision set to 500KHz bandwidth.
a Note: Amplifiers which use bootstrap resistors and capactor as the VAS load, bias setting is impossible as the VAS current varies with rail voltage.
Depending on the amplifier I am working on a compromise sometimes has to be made with [1] and [2] above depending on the VAS stage design, bias spreader proximity to the output devices and the output stage design/values.
I must confess that Zed built 10s of thousands of Hifonics CUPID amplifiers in the late 80s and early 90s - using the dreaded TIP142/147 darlington devices but I did use 1 ohm emitter resistors each with a high speed bypass diode and the bias spreaderdevice was very close to the power devices. We were able to bias these reasonably hard without runaway. I repair Hifonics amplifiers and have never repaired a CUPID.
One of the big drawbacks of using these darlington devices is that they have base-emitter resistors built in and these vary in value from about 8K ohm to 4K ohm and so the VAS stage must drive the output rail through these resistors. I have not seen any designs where a pre-pre driver is used between the VAS and the darlington as the whole idea of using the darlington is to save extra devices.
Wade you wrote:
You mentioned that the SOA on the TIP102 at 30v is 2.7A it’s actually more like 2.8A,
Call me a liar for 0.1A at 40v it’s 2A according to chart (pg. 1). Then you said that the SOA of the TIP142 at 30v is 4A it’s not, on the chart (pg. 2) I provided you will see its only 3A, you say at 40v its 2.8A but it’s actually 1.7A, that’s a 15% increase with the 102’s. If you look at the last chart (pg. 3) that shows the TIP102 at 100ms, you will see @ 100ms at 30v the tIP102 has a SOA of 3.8A. Chart 1 and 2 were downloaded yesterday from Fairchild. Chart 3 is from Motorola. I do not work with 100ms SOA charts, only DC.
To Wade. I am pasting the SOA curve of the TIP devices below.
At 30v VCE I see 4A and at 40v I see about 2.7A. Hope my eyesight is not playing tricks on me.
Best regards,
Steve Mantz.
Special mention to Perry Babin for all the good work he has done on this forum.
Secondly to ALL members on this forum, I value all of you, newbies to the most technical. You all make it worthwhile.
Please understand that this reply is not bashing anyone, just my perspective as a design engineer and manufacturer of amplifiers.
Wade it is good to meet you at last. I tried to purchase Soundstream when they were on their way out but my bid was not accepted.
I can only write what I have experienced with SS amplifiers. I look at amplifiers through an engineer's lense.
The orginal D series in my opinion were not very good amplifiers. A self oscillating supply using TIP35 devices is/was not my first choice of power supply design. There is ZERO dead time in these supplies and running at 20KHz is also not optimum. The amplifiers used TO218 darlingtons [TIP142/147] pairs with 0.1 ohm emitter resistors. These and all other plastic darlingtons have big thermal issues and if biased a little too hard, they will run away thermally.
Reference series were a step forward using a SG3525 as the SMPS control chip but the supplies still had issues. I have experienced transformer failures as Perry pointed out. The largest parts failure I have experienced is the rectifier diodes [FEP16/FEN16] where they are paralleled. It is a golden rule not to parallel silicon diodes withour current sharing resistors in the AC legs. Solution use TO247 diodes and use just one pair! Of course the darlington output devices still fail.
Did I mention that to repair these Reference amplifiers one must remove a zillion screws, carefully remove the PCB from the chassis (Oh those with the muliple LEDs to put the PCB back into the chassis), fix the fault and go through the same gymnastics to put the PCB back into the chassis. A time consuming process!
The amplifiers all suffered the same problems that the darlingtons are well know for. To SSs credit they did eventually change to 1 ohm + diode.Without a high value emitter resistor [1 ohm] and a high speed bypass diode, the devices WILL fail if the amplifier's bias spreader voltage is anywhere close to the 2.25-2.35v. NOTE: I always set idle current using three methods.
1] I keep a volt meter across the bias spreader.
2] I drive the amplifier at 20KHz at about 0.5v RMS into a 2 ohm load
3] I look at the THD on my Audio Precision set to 500KHz bandwidth.
a Note: Amplifiers which use bootstrap resistors and capactor as the VAS load, bias setting is impossible as the VAS current varies with rail voltage.
Depending on the amplifier I am working on a compromise sometimes has to be made with [1] and [2] above depending on the VAS stage design, bias spreader proximity to the output devices and the output stage design/values.
I must confess that Zed built 10s of thousands of Hifonics CUPID amplifiers in the late 80s and early 90s - using the dreaded TIP142/147 darlington devices but I did use 1 ohm emitter resistors each with a high speed bypass diode and the bias spreaderdevice was very close to the power devices. We were able to bias these reasonably hard without runaway. I repair Hifonics amplifiers and have never repaired a CUPID.
One of the big drawbacks of using these darlington devices is that they have base-emitter resistors built in and these vary in value from about 8K ohm to 4K ohm and so the VAS stage must drive the output rail through these resistors. I have not seen any designs where a pre-pre driver is used between the VAS and the darlington as the whole idea of using the darlington is to save extra devices.
Wade you wrote:
You mentioned that the SOA on the TIP102 at 30v is 2.7A it’s actually more like 2.8A,
Call me a liar for 0.1A at 40v it’s 2A according to chart (pg. 1). Then you said that the SOA of the TIP142 at 30v is 4A it’s not, on the chart (pg. 2) I provided you will see its only 3A, you say at 40v its 2.8A but it’s actually 1.7A, that’s a 15% increase with the 102’s. If you look at the last chart (pg. 3) that shows the TIP102 at 100ms, you will see @ 100ms at 30v the tIP102 has a SOA of 3.8A. Chart 1 and 2 were downloaded yesterday from Fairchild. Chart 3 is from Motorola. I do not work with 100ms SOA charts, only DC.To Wade. I am pasting the SOA curve of the TIP devices below.
At 30v VCE I see 4A and at 40v I see about 2.7A. Hope my eyesight is not playing tricks on me.
Best regards,
Steve Mantz.
Special mention to Perry Babin for all the good work he has done on this forum.
You're a real gem, Stephen. I'm glad you and your family are well and navigating this particularly shitty time in human history.
Wade passed away in 2021. So he won't be able to answer your retort. I'm sure he's yelling at you from wherever he is now and i'm sure he didn't take the engineering banter personally. It's in an engineers nature to argue with other engineers. Even more so is their nature to argue with the techs. It's just the way of things and a necessary competition of though that brings about the best results, eventually.
DIY audio wouldn't be the place that it is without all these brilliant minds collectively lobbying their ideas and comparing notes openly. For that I am thankful.
Be well.
MatthewS,
Thanks for your kind words.
I have repaired pretty much every style of Soundstream amplifier. Of course in later years they switvhed to conventional BJTs in the output stages.
We used Darlington power devices in two Hifonics amps we built, we built about 47,000 of them, 2 and 4 channel but of course I used 1 ohm emitter resistors with bypass Shottky diodes. These were the Cupiod and Juno.
Since then I have steered well away from these nasty devices.
Regards
Steve Mantz
Thanks for your kind words.
I have repaired pretty much every style of Soundstream amplifier. Of course in later years they switvhed to conventional BJTs in the output stages.
We used Darlington power devices in two Hifonics amps we built, we built about 47,000 of them, 2 and 4 channel but of course I used 1 ohm emitter resistors with bypass Shottky diodes. These were the Cupiod and Juno.
Since then I have steered well away from these nasty devices.
Regards
Steve Mantz
Wade will be missed. One of his employees gave me a bunch of FEB and other components for Reference series and also gave me a engineering sample of a Reference 405. It's interesting that they crammed 2 pair of finals for each of the 4 channels instead of the production 1 pair per channel. It's never been in a car and maybe has an hour of music played through it. I don't know what to do with it maybe sell it.
Scott.
Scott.