How to choose an isolation transformer to go between op amp and power amp? I am using a 600-600 transformer from an old modem.( Here we go again, try parts and watch for smoke!) It kills the hum but the bandwidth is 300- 4k. looking at jensons but the price is high. found a triad ty250p 20hz to20khz and a variety of impedances to choose from how does a person determine the correct impedance to order? Also how to avoid saturation,( hear thats a bad thing) when driving the tranny from the op amp output? Any help would be appreciated🙂
Presumably the transformer primary should offer a high impedance to the op amp and the transformer secondary should offer a low impedance to the power amp, but I admit to not being sure about your application.
More expensive transformers are less likely to saturate.
More expensive transformers are less likely to saturate.
Thanks for your response Galu. My power amp has an input impedance of 55k at 1 watt 1kohm test signal Should i be trying to match that? and how do i determine how much output the op amps can dump into the tranny? there must be a formula or rule of thumb?
Hi Jimmy!
You have not explained your set up in any detail, but I assume that the only reason for using the transformer is to reduce hum.
When using a transformer, I would ensure that the primary impedance is around ten times the output impedance of the op amp and that the secondary impedance is around 1/10th of the input impedance of the power amp.
A Triad TY-142P (10k/2k) may be more suitable than your present transformer, provided its frequency range of 200Hz to 15,000Hz +/- 2dB proves adequate.
Unfortunately, audio transformers with a better frequency range are substantially more expensive than the Triad.
Regarding 'how much output', you will be aware that op amps are not capable of delivering high currents.
Therefore the rule of thumb is that the tranny primary impedance should be around 10 times higher than the op amp output impedance.
This is the rule I have tried to apply in choosing your transformer, although I obviously don't know the exact output impedance of your op amp.
You have not explained your set up in any detail, but I assume that the only reason for using the transformer is to reduce hum.
When using a transformer, I would ensure that the primary impedance is around ten times the output impedance of the op amp and that the secondary impedance is around 1/10th of the input impedance of the power amp.
A Triad TY-142P (10k/2k) may be more suitable than your present transformer, provided its frequency range of 200Hz to 15,000Hz +/- 2dB proves adequate.
Unfortunately, audio transformers with a better frequency range are substantially more expensive than the Triad.
Regarding 'how much output', you will be aware that op amps are not capable of delivering high currents.
Therefore the rule of thumb is that the tranny primary impedance should be around 10 times higher than the op amp output impedance.
This is the rule I have tried to apply in choosing your transformer, although I obviously don't know the exact output impedance of your op amp.
Last edited:
Low frequency cutoff should also be considered. For that you need to know the inductance of the primary. The formula is f = R / (2 * PI * L)
Transformers have often somewhat limited banwidth.
There are more expensive audio ones, which have upper frequency 20k.
Why do you use isolationtransformer?
Better to do without.
There are more expensive audio ones, which have upper frequency 20k.
Why do you use isolationtransformer?
Better to do without.
Taobao has one or two highly affordable transformer manufacturers, there's no real need to spend a fortune even for a high nickel content core. Check out these - a pair is comfortably under $15 but you'll need to find a Taobao agent most likely.
????10K:10K ??????? ???80%???????-???
????10K:10K ??????? ???80%???????-???
Hi Jimmy!
If there's a Triad 10k/10k with a wider frequency range then that should do the job.
Here's more information on saturation:
The op amp could have dc on the output. This would produce a permanent magnetic field in the transformer.
This 'dc field' could be large enough to cause the transformer windings to become magnetically saturated. This would distort the ac signal which is being transformed.
This is avoided by connecting a blocking capacitor in series with the transformer primary. The capacitor should be non polar and in the hundreds of microfarad range.
This may be achieved by connecting two polar 100uF/25V electrolytics in series but in reverse i.e. with the two negative terminals of the capacitors connected together in the middle. The combination may be bypassed by a 0.1uF film capacitor.
Hope this helps!
If there's a Triad 10k/10k with a wider frequency range then that should do the job.
Here's more information on saturation:
The op amp could have dc on the output. This would produce a permanent magnetic field in the transformer.
This 'dc field' could be large enough to cause the transformer windings to become magnetically saturated. This would distort the ac signal which is being transformed.
This is avoided by connecting a blocking capacitor in series with the transformer primary. The capacitor should be non polar and in the hundreds of microfarad range.
This may be achieved by connecting two polar 100uF/25V electrolytics in series but in reverse i.e. with the two negative terminals of the capacitors connected together in the middle. The combination may be bypassed by a 0.1uF film capacitor.
Hope this helps!
Not sure what you mean by an "op amp" (is that your term for a preamp?) but I'm curious - why do you think an isotran will kill the hum you are hearing?
Andy
When he says "op amp" he probably means 'op amp' i.e. an operational amplifier, probably on a chip. This op amp may be used a line stage or preamp; op amp is a component, while preamp is an application.
Hi Andy!
I was unsure about jimmy audio's objective from the start.
You may be correct that he meant pre amp instead of op amp.
I hope he will re-engage with his thread to clarify matters.
I was unsure about jimmy audio's objective from the start.
You may be correct that he meant pre amp instead of op amp.
I hope he will re-engage with his thread to clarify matters.
I use the Triad TY-146 often. Load it on the 600 ohm output with a 680 ohm resistor 1/4 watt in series with a .022 uf film capacitor. Use the input as 600 ohms. As you will not be using it at 1000 mW the low frequency range will extend below 20 hertz.
I agree with loading the tansformer secondary.
If using a 10K/10k transformer the loading would be 10k in series with 0.001uF.
If using a 10K/10k transformer the loading would be 10k in series with 0.001uF.
Exactly - which is why I find it strange to want to put an opamp straight in front of a power amp ... you need all the other bits which turn it (the opamp) into an application.
Andy
Sorry it took so long to reply, work kept me in the land without internet for the past week. My project has a 24 volt single ended supply for the speaker protect board. a single ended 12 volt supply for the remote receiver. a 15 volt single ended supply for the mains power relay and latch circuit. a + 12 -12 split supply for the summing amps/ volume control, a 24 volt supply built in to the soft start board. also 2 +65 -65 supplies for the six amp boards (5.1 configuration using stk 4048 chips) While no trannys in the signal path would be best case scenario with all the supplies and a rf receiver for remote noise is substantial and all my efforts to reduce or eliminate it ( star grounding, lots of filter capacitance careful routing of shielded input cables ect have largely failed. So adding trannys at the power amp input became the fix. They work fine other than the reduced bandwith due to crappy trannys, the noise floor is dead silent and stays that way as I run the volume all the way up( 1.26 mv measured worst case. So the reason for my post was to get a feel for what kind of trannys would be appropriate. I,m guessing the Jensons should be fairly transparent but at $70-80 a piece that adds up quick Edcores look good but 8 weeks out. Triad Ty-250p +- 1db from 20 to 20k, 10 k primary and secondary looks like a good bet( only $6) and available. Anyone know of anything better? would it be worth the wait for the edcores? and most importantly should I add a cap to kill any dc from the op amps from stacking up the trannys? if so would a small electrolytic do the trick 1- 4.7 uf?
Are we correct in assuming that your project is a preamp then?
Electrolytic capacitor? Is the info in post #9 any use?
See also post #14.
Worth trying the affordable Triads before committing large sums of money to Jensons. Only by trying them out will you be able to establish if the frequency range is wide enough for your application.
Hey Galu, My project is a 5.1 amp. It's connected to my flatscreen via Toslink to a hd audio rush decoder. from there 6 rca cables to the amp. For volume control I hacked a yahama volume control board with alps 6 gang motorized pot and h bridge. interfaces with my remote nicely. Thats where the op amps come in jrs 2068 used in an active volume control arrangement with 5.4 db gain on the front r and l speakers 26db on the rear speakers a little more gain on the sub and center 28.9 i believe.... post 9, a 10k and.001 across the trannys secondary? what about dc on the primary?
The 10k & 0.001uF in series across the secondary damps the secondary resonance at high frequencies at the expense of high frequency loss above approximately 16kHz. Use this if there's any hint of oscillation.
dc on the primary is dealt with in post #9.
- Status
- Not open for further replies.