TA7642 circuit voltage supply - UK Vintage Radio Repair and Restoration Discussion Forum

Jolly 7 · 15th Nov 2019, 2:47 pm

I found this circuit on RS electronics website, all credits to Chelsea Back.
The circuit's designed to operate on 9V, but works perfectly on 1.5V. Can someone please explain why it functions at such a low voltage and is it safe to go up to 9V ? The max tolerated voltage for TA7642 seems to be only 6V.

ionburn · 15th Nov 2019, 3:02 pm

The IC will never see 9V with the diodes in circuit as they are effectively regulators. It is designed to work on low voltage:

http://www.shenzhensum.com/products/...eet/TA7642.pdf

Skywave · 15th Nov 2019, 3:14 pm

Quote:

Originally Posted by Jolly 7

The circuit's designed to operate on 9V.
Is it safe to go up to 9V ? The max tolerated voltage for TA7642 seems to be only 6V.

Yes, the circuit is powered from 9v., but the i.c. isn't. Hence the two diodes in the supply rail.
A supply of 9v. will be O.K., provided that the current through the diodes is not in excess of their rating and their series resistor has a value to accommodate that requirement.
The reason it operates at such a low voltage is simply because the i.c. has been designed for that low voltage.

Al.

Jolly 7 · 15th Nov 2019, 7:04 pm

Thanks for the responses. The diodes being used in this circuit are IN4148s. Could IN4007s also be used without detriment ?

G6Tanuki · 15th Nov 2019, 7:07 pm

Yes, exactly as Skywave says, the two forward-biased diodes act as a voltage-regulator based on the typical 0.6V conduction-threshold of a non-Schottky silicon P-N junction - so two forward-biased Si diodes give a 'stabilised' voltage of around 1.2 or 1.3V for the ZN414-clone IC.

When you're only supplying a couple of Volts of battery-power to that circuit the two diodes probably won't be seeing enough voltage to really overcome their conduction-threshold. But when the supply is 9V the diodes will be 'doing their thing' and holding the voltage down.

Neil Purling · 15th Nov 2019, 8:13 pm

I made a circuit from 'Electronics Australia'. I eliminated the amplification stages.
The attraction of it was an added RF Gain control and that it operated from 1.5 volts. If that is what the circuit needs to operate the diodes are wasting 5/6 of the battery's voltage.
The RF Gain was a 100K Lin pot with the track between the o/p lead of the TA7642 & chassis. The wire of the 100K fixed resistor soldered to the same spot is removed and soldered to the pot's wiper connection.

Jolly 7 · 15th Nov 2019, 8:56 pm

I find that I can get away without using a gain control circuit in my location just outside of Southampton, but that's maybe because there are no strong MW transmitters nearby that could swamp out weaker stations and occupy parts of the dial where they're not meant to be. Using the circuit in my first post, I can get full MW coverage. At the moment I have it working inside an old broken KB Rhapsody valve radio, with its valves and other components already removed.

Jolly 7 · 16th Nov 2019, 3:11 am

Having attempted to run this circuit higher than 1.5V, I found that distortion set in around 2V and around 4.5V so much noise appeared that the radio became unusable. Not sure if the circuit can be modified to overcome this problem ?

Argus25 · 16th Nov 2019, 4:20 am

The output of the amplifier , with its 1k load resistor is clobbered with a 0.1uF capacitor, That will eliminate any amplified RF and muffle any audio frequencies over a few kHz.

Looking at the internal diagram, there is a high Z input stage with a current mirror and then some amplification stages. I cannot see a proper detector so I assume they rely on the non linearity towards the output stage to demodulate the carrier. Even so frequencies above 1.5kHz will be 3 dB down. You might want to increase that 0.1uF cap to 0.022uF or something like that.

I'm not really surprised about the distortion.

Also, the bypass capacitor for the DC feedback would be better as a 0.1uF rather than a 0.01uF. In fact I wonder if they fouled up the data sheet and is supposed to have a 0.01uF on the output and a 0.1 or 0.15uF as the bybass cap on the 100k feedback resistor on the input circuit. Because with their 1.5K load resistor (data sheet example), and a 0.01uF on the output, it would start to roll off at about 10.6kHz or at about 16kHz with a 1k load resistor.

The reason it can run off a low voltage is they really only need to apply enough voltage to overcome the transistor's b-e drop and a little more to get some bias current. So circuits with grounded emitters will run to low supply voltages close to 1 to 1.5V, but to DC stabilize them they have to acquire some of the base bias from the collector circuits, so there is at least some DC negative feedback. It was a common method in hearing aids running off 1.5V batteries to ground the emitters and use this bias methodology.

Argus25 · 16th Nov 2019, 9:51 pm

Actually I think the way this IC works to detect, is that the audio at the output terminal modulates the gain of the preceding stages as they are powered by the output terminal, so the detection is due to square law amplification of the carrier or similar. As the output voltage increases, so does the gain.

Jolly 7 · 16th Nov 2019, 10:03 pm

Thanks for the detailed explanation. When you say to try and increase the output cap from 0.1 to 0.022 uF, do you actually mean 0.22 uF ? Has an extra zero crept in ?

Argus25 · 16th Nov 2019, 10:09 pm

Quote:

Originally Posted by Jolly 7

Thanks for the detailed explanation. When you say to try and increase the output cap from 0.1 to 0.022 uF, do you actually mean 0.22 uF ? Has an extra zero crept in ?

sorry, just a typo, should say decrease.

Jolly 7 · 16th Nov 2019, 10:37 pm

No worries. Here is another TA7642 circuit I deployed in my first TA7642 radio. This works well and unlike the 'Chelsea Black' circuit in my first post, only using the primary winding of the antenna coil gives very good results. Connecting the secondary winding of the coil led to missing stations, stations appearing on incorrect parts of the dial or even duplicate stations. The gain of this circuit is consistent across 531 to 1557 kHz.
Corrections and improvement suggestions welcome. I would like to add an audio amp stage using one or two transistors but not sure how to get about it.

Argus25 · 16th Nov 2019, 11:50 pm

Quote:

Originally Posted by Jolly 7

No worries. Here is another TA7642 circuit I deployed in my first TA7642 radio. This works well and unlike the 'Chelsea Black' circuit in my first post, only using the primary winding of the antenna coil gives very good results. Connecting the secondary winding of the coil led to missing stations, stations appearing on incorrect parts of the dial or even duplicate stations. The gain of this circuit is consistent across 531 to 1557 kHz.
Corrections and improvement suggestions welcome. I would like to add an audio amp stage using one or two transistors but not sure how to get about it.

It looks ok. I'm surprised it has such heavy output filtering with three 0.1uF caps.

In any case if I made it, I would transmit to it from a distance with a test RF generator with variable frequency audio modulation, to find out when the upper audio frequencies rolled off. It might be that the output Z of the IC is low enough that the load capacitance there does not roll it off as much as it appears it could.

Argus25 · 17th Nov 2019, 1:16 am

Jolly_7,

To be frank when I saw your IC it reminded me of the Shania Twain song "That don't impress me much"

If you are going down the road of a single IC TRF radio, you could go for something with some real class like the LM372, the internal architecture of the IC is far superior and there is the output stage design in the project you require to drive a speaker. And make a proper job of it too with the construction.

I have attached an article from Electronics Australia, 1969. If you make this radio, not only will you get good results, but you will have a lovely little radio to impress your friends with too.

If the images are too difficult here, I will email you the scans. Part of the fun would be hunting this IC down.

Also, just to advise you, this IC also came in an epoxy package as well, here is a link to one just to show the photo of what this variant was like back in 1969. (not to promote the seller or anything)

https://www.ebay.com/sch/i.html?_fro...372CN&_sacat=0

Sorry I messed up the link somehow, this is the direct link:

https://www.ebay.com/itm/LM372CN-NSC...AAAOSwHLNZQrpa

Hugo.

Jolly 7 · 17th Nov 2019, 2:15 am

Many thanks Hugo. Most of the jpegs are legible apart from the very edges of some of the pages. I will let you know if I need more info. I would love to build the radio if I can get hold of the IC. I wasn't aware of its existence. Would be nice if they still had it in production.

Argus25 · 17th Nov 2019, 2:16 am

......PS : I hope the remark about your IC doesn't lead me to incur the wrath of the ZN414 enthusiasts, as your IC is more or less a clone of it. And these IC's do have the advantage of a very high input Z over the LM372 and a little more gain. But it is not really a problem as the 372 can simply be driven by an existing coil tap or small coil normally present on the standard ferrite rod coil assembly.

I also saw some LM372's for sale on AU ebay, but they seemed a little costly, but you could always make an offer to the seller.

In some ways I find it more fun to build things with rare and obsolete parts, than common garden ones, it makes it special somehow.

15th Nov 2019, 2:47 pm	#1
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	TA7642 circuit voltage supply I found this circuit on RS electronics website, all credits to Chelsea Back. The circuit's designed to operate on 9V, but works perfectly on 1.5V. Can someone please explain why it functions at such a low voltage and is it safe to go up to 9V ? The max tolerated voltage for TA7642 seems to be only 6V. Attached Thumbnails

15th Nov 2019, 3:02 pm	#2
ionburn Heptode Join Date: Dec 2014 Location: Lincoln, Lincolnshire, UK. Posts: 583	Re: TA7642 circuit voltage supply The IC will never see 9V with the diodes in circuit as they are effectively regulators. It is designed to work on low voltage: http://www.shenzhensum.com/products/...eet/TA7642.pdf

15th Nov 2019, 7:04 pm	#4
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: TA7642 circuit voltage supply Thanks for the responses. The diodes being used in this circuit are IN4148s. Could IN4007s also be used without detriment ?

15th Nov 2019, 7:07 pm	#5
G6Tanuki Dekatron Join Date: Apr 2012 Location: Wiltshire, UK. Posts: 13,996	Re: TA7642 circuit voltage supply Yes, exactly as Skywave says, the two forward-biased diodes act as a voltage-regulator based on the typical 0.6V conduction-threshold of a non-Schottky silicon P-N junction - so two forward-biased Si diodes give a 'stabilised' voltage of around 1.2 or 1.3V for the ZN414-clone IC. When you're only supplying a couple of Volts of battery-power to that circuit the two diodes probably won't be seeing enough voltage to really overcome their conduction-threshold. But when the supply is 9V the diodes will be 'doing their thing' and holding the voltage down.

15th Nov 2019, 8:13 pm	#6
Neil Purling Nonode Join Date: Jul 2005 Location: Hull, East Yorkshire, UK. Posts: 2,086	Re: TA7642 circuit voltage supply I made a circuit from 'Electronics Australia'. I eliminated the amplification stages. The attraction of it was an added RF Gain control and that it operated from 1.5 volts. If that is what the circuit needs to operate the diodes are wasting 5/6 of the battery's voltage. The RF Gain was a 100K Lin pot with the track between the o/p lead of the TA7642 & chassis. The wire of the 100K fixed resistor soldered to the same spot is removed and soldered to the pot's wiper connection.

15th Nov 2019, 8:56 pm	#7
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: TA7642 circuit voltage supply I find that I can get away without using a gain control circuit in my location just outside of Southampton, but that's maybe because there are no strong MW transmitters nearby that could swamp out weaker stations and occupy parts of the dial where they're not meant to be. Using the circuit in my first post, I can get full MW coverage. At the moment I have it working inside an old broken KB Rhapsody valve radio, with its valves and other components already removed.

16th Nov 2019, 3:11 am	#8
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: TA7642 circuit voltage supply Having attempted to run this circuit higher than 1.5V, I found that distortion set in around 2V and around 4.5V so much noise appeared that the radio became unusable. Not sure if the circuit can be modified to overcome this problem ?

16th Nov 2019, 4:20 am	#9
Argus25 No Longer a Member Join Date: Oct 2016 Location: Maroochydore, Queensland, Australia. Posts: 2,679	Re: TA7642 circuit voltage supply The output of the amplifier , with its 1k load resistor is clobbered with a 0.1uF capacitor, That will eliminate any amplified RF and muffle any audio frequencies over a few kHz. Looking at the internal diagram, there is a high Z input stage with a current mirror and then some amplification stages. I cannot see a proper detector so I assume they rely on the non linearity towards the output stage to demodulate the carrier. Even so frequencies above 1.5kHz will be 3 dB down. You might want to increase that 0.1uF cap to 0.022uF or something like that. I'm not really surprised about the distortion. Also, the bypass capacitor for the DC feedback would be better as a 0.1uF rather than a 0.01uF. In fact I wonder if they fouled up the data sheet and is supposed to have a 0.01uF on the output and a 0.1 or 0.15uF as the bybass cap on the 100k feedback resistor on the input circuit. Because with their 1.5K load resistor (data sheet example), and a 0.01uF on the output, it would start to roll off at about 10.6kHz or at about 16kHz with a 1k load resistor. The reason it can run off a low voltage is they really only need to apply enough voltage to overcome the transistor's b-e drop and a little more to get some bias current. So circuits with grounded emitters will run to low supply voltages close to 1 to 1.5V, but to DC stabilize them they have to acquire some of the base bias from the collector circuits, so there is at least some DC negative feedback. It was a common method in hearing aids running off 1.5V batteries to ground the emitters and use this bias methodology. Last edited by Argus25; 16th Nov 2019 at 4:31 am.

16th Nov 2019, 9:51 pm	#10
Argus25 No Longer a Member Join Date: Oct 2016 Location: Maroochydore, Queensland, Australia. Posts: 2,679	Re: TA7642 circuit voltage supply Actually I think the way this IC works to detect, is that the audio at the output terminal modulates the gain of the preceding stages as they are powered by the output terminal, so the detection is due to square law amplification of the carrier or similar. As the output voltage increases, so does the gain.

16th Nov 2019, 10:03 pm	#11
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: TA7642 circuit voltage supply Thanks for the detailed explanation. When you say to try and increase the output cap from 0.1 to 0.022 uF, do you actually mean 0.22 uF ? Has an extra zero crept in ?

16th Nov 2019, 10:37 pm	#13
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: TA7642 circuit voltage supply No worries. Here is another TA7642 circuit I deployed in my first TA7642 radio. This works well and unlike the 'Chelsea Black' circuit in my first post, only using the primary winding of the antenna coil gives very good results. Connecting the secondary winding of the coil led to missing stations, stations appearing on incorrect parts of the dial or even duplicate stations. The gain of this circuit is consistent across 531 to 1557 kHz. Corrections and improvement suggestions welcome. I would like to add an audio amp stage using one or two transistors but not sure how to get about it. Attached Thumbnails

17th Nov 2019, 1:16 am	#15
Argus25 No Longer a Member Join Date: Oct 2016 Location: Maroochydore, Queensland, Australia. Posts: 2,679	Re: TA7642 circuit voltage supply Jolly_7, To be frank when I saw your IC it reminded me of the Shania Twain song "That don't impress me much" If you are going down the road of a single IC TRF radio, you could go for something with some real class like the LM372, the internal architecture of the IC is far superior and there is the output stage design in the project you require to drive a speaker. And make a proper job of it too with the construction. I have attached an article from Electronics Australia, 1969. If you make this radio, not only will you get good results, but you will have a lovely little radio to impress your friends with too. If the images are too difficult here, I will email you the scans. Part of the fun would be hunting this IC down. Also, just to advise you, this IC also came in an epoxy package as well, here is a link to one just to show the photo of what this variant was like back in 1969. (not to promote the seller or anything) https://www.ebay.com/sch/i.html?_fro...372CN&_sacat=0 Sorry I messed up the link somehow, this is the direct link: https://www.ebay.com/itm/LM372CN-NSC...AAAOSwHLNZQrpa Hugo. Attached Thumbnails Last edited by Argus25; 17th Nov 2019 at 1:35 am.

17th Nov 2019, 2:15 am	#16
Jolly 7 Octode Join Date: Feb 2018 Location: Southampton, Hampshire, UK. Posts: 1,061	Re: TA7642 circuit voltage supply Many thanks Hugo. Most of the jpegs are legible apart from the very edges of some of the pages. I will let you know if I need more info. I would love to build the radio if I can get hold of the IC. I wasn't aware of its existence. Would be nice if they still had it in production.

17th Nov 2019, 2:16 am	#17
Argus25 No Longer a Member Join Date: Oct 2016 Location: Maroochydore, Queensland, Australia. Posts: 2,679	Re: TA7642 circuit voltage supply ......PS : I hope the remark about your IC doesn't lead me to incur the wrath of the ZN414 enthusiasts, as your IC is more or less a clone of it. And these IC's do have the advantage of a very high input Z over the LM372 and a little more gain. But it is not really a problem as the 372 can simply be driven by an existing coil tap or small coil normally present on the standard ferrite rod coil assembly. I also saw some LM372's for sale on AU ebay, but they seemed a little costly, but you could always make an offer to the seller. In some ways I find it more fun to build things with rare and obsolete parts, than common garden ones, it makes it special somehow.