Just came across the excellent JLH discreet stereo FM decoder article in ETI Feb.'87 & would like further info re later mods. & further articles etc. on this unit.
With a quoted Stereo THD of just 0.02% over 30 - 15Khz. it appears superior to anything available to-day.
I don't suppose anyone is still supplying this as a kit but would appreciate it if anyone can advise on any further information on this subject. - I know that Hart electronics in Shropshire used to do them but they are no longer in business. I did try the williamshart site but it is no longer viable either.
Thanks.
With a quoted Stereo THD of just 0.02% over 30 - 15Khz. it appears superior to anything available to-day.
I don't suppose anyone is still supplying this as a kit but would appreciate it if anyone can advise on any further information on this subject. - I know that Hart electronics in Shropshire used to do them but they are no longer in business. I did try the williamshart site but it is no longer viable either.
Thanks.
You may find information about this decoder searching in this forum.
I have built the Hart kit. A real disappointment.
The THD was claimed low but the diapahony was really too large.
As far as I remember, it has been said that there was something insufficiently thought in the concept of this Sample and Hold decoder.
I can provide the JLH articles.
I have built the Hart kit. A real disappointment.
The THD was claimed low but the diapahony was really too large.
As far as I remember, it has been said that there was something insufficiently thought in the concept of this Sample and Hold decoder.
I can provide the JLH articles.
Last edited:
If that is the decoder which samples at the peaks of the 38kHz waveform then I think it is based on a confusion (or actually two confusions).
The first is that sampling somehow reduces noise. If anything it will increase noise, as the sampling waveform will pick up noise around all harmonics of the sampling frequency; unlike a switching decoder which only gets noise around the odd harmonics and that decreasing with harmonic order.
The second is that sample and hold suffers from the usual sinc frequency response (e.g. as seen in NOS DACs, but here it is slightly worse as the sampling frequency is lower). I don't recall that he applied any correction for this so it will have an HF droop.
A switching decoder is better, provided that the necessary crosstalk is included to allow for the difference in fundamental amplitude between a sine wave and a square wave.
JLH was a popular designer, but he had some gaps in his knowledge which he was very reluctant to admit.
The first is that sampling somehow reduces noise. If anything it will increase noise, as the sampling waveform will pick up noise around all harmonics of the sampling frequency; unlike a switching decoder which only gets noise around the odd harmonics and that decreasing with harmonic order.
The second is that sample and hold suffers from the usual sinc frequency response (e.g. as seen in NOS DACs, but here it is slightly worse as the sampling frequency is lower). I don't recall that he applied any correction for this so it will have an HF droop.
A switching decoder is better, provided that the necessary crosstalk is included to allow for the difference in fundamental amplitude between a sine wave and a square wave.
JLH was a popular designer, but he had some gaps in his knowledge which he was very reluctant to admit.
Hi
Order this and be done with it! Use it to replace decoder in almost any tuner for an improvement. Or to add stereo to a mono tube unit.
Implementation
But best ever ....well find a sanyo la3450 chip and build using the datasheet. well only a few difficult options could do better still.
Cheers
Paba
Order this and be done with it! Use it to replace decoder in almost any tuner for an improvement. Or to add stereo to a mono tube unit.
Implementation
But best ever ....well find a sanyo la3450 chip and build using the datasheet. well only a few difficult options could do better still.
Cheers
Paba
That is what I use. My DIY FM tuner has
Dual-gate MOSFET RF amp, preceded by PIN diode attenuator
Schottky diode balanced mixer (fed from JFET local osc and MOSFET buffer amp)
JFET IF preamp
Ceramic filter
BJT LTP IF amp
Ceramic filter (diode switched for bandwidth)
Hitachi IF/discriminator chip (with dual Toko coil)
Hitachi stereo decoder chip
The RF front end uses techniques more commonly seen in a communications receiver. The back end was largely based on data from Ambit International catalogues. The amplified AFC has its loop gain adjusted according to signal strength.
Dual-gate MOSFET RF amp, preceded by PIN diode attenuator
Schottky diode balanced mixer (fed from JFET local osc and MOSFET buffer amp)
JFET IF preamp
Ceramic filter
BJT LTP IF amp
Ceramic filter (diode switched for bandwidth)
Hitachi IF/discriminator chip (with dual Toko coil)
Hitachi stereo decoder chip
The RF front end uses techniques more commonly seen in a communications receiver. The back end was largely based on data from Ambit International catalogues. The amplified AFC has its loop gain adjusted according to signal strength.
It isn't much to look at, just a metal box with black painted front panel. The circuit diagrams are still in a box somewhere in the house. Some years ago I was thinking of writing it up for a magazine article (probably Wireless World) but then DAB came along and we were told FM would not last very long in the UK.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.