First, forgive me that I have not posted here in ages. I'm looking for help with my much loved Threshold FET ten/hl preamp. In my system the outputs of the FET ten/hl are connected to a homebrew converter using Jensen JT-6110K-B 4:1 stepdown line input transformers. The balanced outputs from my transformer box feed a pair of Schiit Tyr monoblocks.
The problem is I have been measuring about a 5 dB drop off from 40 Hz to 20,000 Hz at the outputs of the Tyrs.. My first thought was the problem might be caused by excess capacitance in the cables running from the transformer box. I swapped out cables. I removed the transformer box. I swapped out a lot of things. The only thing that fixed the drop off problem was disconnecting the FET ten/hl and putting in a different preamp. Nothing else made any difference.
Would anyone have thoughts or suggestions?
The problem is I have been measuring about a 5 dB drop off from 40 Hz to 20,000 Hz at the outputs of the Tyrs.. My first thought was the problem might be caused by excess capacitance in the cables running from the transformer box. I swapped out cables. I removed the transformer box. I swapped out a lot of things. The only thing that fixed the drop off problem was disconnecting the FET ten/hl and putting in a different preamp. Nothing else made any difference.
Would anyone have thoughts or suggestions?
Sure, but it can drive a 10k load as well. You'll find a perfectly flat response with any load from 10k to open circuit.
Did you try it directly with the Schiit Tyr in single ended input mode?
Did you try it directly with the Schiit Tyr in single ended input mode?
I am curious what the input source is on the Threshold? A sweep generator? Vinyl record test tone? And, you suggest a 5db drop from 40 to 20k. Is that referenced to previous measurements? Is there distortion? If it truly it is a 5 db drop: then maybe really it is 6 and suggests one side of your balanced outputs are missing.
Yes, I did try connecting the Tyr directly in single ended mode. In addition to the roll off there was audible hiss and hum. With the transformers the background is dead silent.
Thanks, I did the experiment, I measured the output of the Threshold with no load. I saw the same drop off as when I measured at the output of the Tyr.
The source of the tones are wave files generated in Audacity. The files are streamed using foobar from my NAS through a HoloAudio Red and converted by a NAD M51 DAC.
The drop is measured from 40 Hz. For example in one test I measured:
40 Hz 2.5182 V
400 Hz 2.4640 V
4000 Hz 2.4105 V
16000 Hz 1.6576 V
20000 Hz 1.3514 V
I have no equipment for measuring distortion, but there is no distortion that I can hear.
What do you use for the voltage measurement, a DVM? Most don't go that high.
Did you normalize the output relative to the input?
Either adjust each input frequency to be at the exact same input level (for example 1.00V), before measuring the output
or else for each input frequency divide the output by the input voltage to get the gain.
Did you normalize the output relative to the input?
Either adjust each input frequency to be at the exact same input level (for example 1.00V), before measuring the output
or else for each input frequency divide the output by the input voltage to get the gain.
Sounds like the power amp is not getting a fully balanced signal. And the noisy single end may tell the story of why. If it is single end all the way to the output and then uses a phase inverter to create the balanced output and that circuit is in trouble it may be loading the single end output in a less than optimal way. I don’t have one or a diagram but have had what you describe trace back to that And no balanced input works very well unless it is being fed a truly balanced signal.
this is crucial
either scope or proper AC wide-range voltmeter
DVM is good to few KHz, no good above
Thanks everyone! My meter is an Agilent 34405A. The response is specified to be +/- 1% at 30kHz.
The noisy single ended connection just sounds like a ground loop to me. I measure the roll off whether I connect the transformer box or drive the amplifiers singled ended.
I did not normalize my outputs relative to the inputs, if I understand correctly. The fact that my little Schiit preamp exhibits no roll off indicates to me that input voltage at each frequency is the same. And I don't see how the input voltages could be different unless the DAC were broken?
Also with the FET ten/hl I measure the same roll off with the load disconnected as I do connected, and measured at the output of the Tyr.
If you think the problem is in the Threshold, you should be measuring it alone, with a standard resistor load of 10k.
Also compare the responses, when the volume control is set at max, and the response with the control set at -6dB.
Has someone replaced the volume control with a different one? Preferred value would be 10k.
Also compare the responses, when the volume control is set at max, and the response with the control set at -6dB.
Has someone replaced the volume control with a different one? Preferred value would be 10k.
I recalled seeing this a few years back and thought perhaps it would help. Particularly page 3 which speaks of a blanket polystyrene cap replacement.
Mystery mainly mostly solved. I have Bose computer speakers connected to one of the tape outputs of FET ten/hl, as mentioned here:
https://www.diyaudio.com/community/threads/fet-ten-hl-recorder-output-puzzle.263453/
Disconnecting the Bose from the FET ten/hl makes a difference in the roll off. These are the measurements taken at the output of the left channel Tyr. FET ten/hl volume control set at the same position for all readings:
With Bose connected
40 Hz 1.036 V
400 Hz 1.014 V
4000 Hz 0.991 V
16000 Hz 0.674 V
20000 Hz 0.547 V
With Bose disconnected
40 Hz 1.314 V
400 Hz 1.316 V
4000 Hz 1.314 V
16000 Hz 1.227 V
20000 Hz 1.152 V
https://www.diyaudio.com/community/threads/fet-ten-hl-recorder-output-puzzle.263453/
Disconnecting the Bose from the FET ten/hl makes a difference in the roll off. These are the measurements taken at the output of the left channel Tyr. FET ten/hl volume control set at the same position for all readings:
With Bose connected
40 Hz 1.036 V
400 Hz 1.014 V
4000 Hz 0.991 V
16000 Hz 0.674 V
20000 Hz 0.547 V
With Bose disconnected
40 Hz 1.314 V
400 Hz 1.316 V
4000 Hz 1.314 V
16000 Hz 1.227 V
20000 Hz 1.152 V
Since my last post I removed a Jenson Iso-Max from between my NAD M51 DAC and the Threshold preamp. Now the response at the output of the left channel amplifier is almost as flat as I can measure:
40 Hz 3.02 V
400 Hz 3.02 V
4000 Hz 3.03 V
16000 Hz 3.07 V
20000 Hz 3.07 V
This is still with my DIY step-down balancing transformer box between the preamp and the power amps. And, don't worry, the Jensen Iso-Max has been put to good use elsewhere.
40 Hz 3.02 V
400 Hz 3.02 V
4000 Hz 3.03 V
16000 Hz 3.07 V
20000 Hz 3.07 V
This is still with my DIY step-down balancing transformer box between the preamp and the power amps. And, don't worry, the Jensen Iso-Max has been put to good use elsewhere.