Poorly issue two Spectrum

User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: Poorly issue two Spectrum

Post by 1024MAK »

Lardo Boffin wrote: Thu Apr 22, 2021 9:40 amThe ULA case gets too hot to comfortably touch after a minute or two. Not sure how quickly that should heat up?
The voltages are all fine. The 9V (nominal) is the input supply voltage, so can be tested on the input power connector. If you only intend to use a modern 9V regulated / SMPSU then there is no need to worry about it.

The ULA: if you mean in the dead centre of the case of the chip, well that’s actually where the chip is. And it’s normal for ULAs to run hot. See the various discussions on this board and on the Spectrum Computing forums about ULAs and heatsinks.

The large resistor (R34) next to the ULA is in series with the main power supply feed to the ULA (pin 13). The voltage here should be between 2.8V to 3.5V. More on this here.

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: Poorly issue two Spectrum

Post by 1024MAK »

More on the insides of a DIL packaged chip here and here ;)

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
Lardo Boffin
Posts: 2160
Joined: Sat Nov 26, 2016 2:42 am

Re: Poorly issue two Spectrum

Post by Lardo Boffin »

Thanks for the info!

The 9V read 11.03V on mine at the regulator pins (once I found an original PSU). Will that vary from PSU to PSU?

Also would fitting a modern voltage regulator, e.g. a Murata, have any effect on this? Aside the being able to get rid of the heat sink that is.
ZX80
ZX81 iss 1 (bugged ROM, kludge fix, normal, rebuilt)
TS 1000 iss 3, ZXPand AY and +, ZX8-CCB, ZX-KDLX & ChromaSCART
Tatung 81 + Wespi
TS 1500 & 2000
Spectrum 16k (iss 1 s/n 862)
Spectrum 48ks plus a DIVMMC future and SPECTRA
Lardo Boffin
Posts: 2160
Joined: Sat Nov 26, 2016 2:42 am

Re: Poorly issue two Spectrum

Post by Lardo Boffin »

The picture is rather good for RF.

CF8865A8-ED48-45A1-870F-7CD0E08A3045.jpeg

But I think the keymat needs attention...
ZX80
ZX81 iss 1 (bugged ROM, kludge fix, normal, rebuilt)
TS 1000 iss 3, ZXPand AY and +, ZX8-CCB, ZX-KDLX & ChromaSCART
Tatung 81 + Wespi
TS 1500 & 2000
Spectrum 16k (iss 1 s/n 862)
Spectrum 48ks plus a DIVMMC future and SPECTRA
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: Poorly issue two Spectrum

Post by 1024MAK »

Lardo Boffin wrote: Thu Apr 22, 2021 1:47 pm The 9V read 11.03V on mine at the regulator pins (once I found an original PSU). Will that vary from PSU to PSU?
Yes, as Sinclair used at least four different suppliers for the UK1400 PSUs. Also as they are unregulated, the output voltage is proportional to the mains voltage. So if the mains voltage is higher, the output will also be higher.
Lardo Boffin wrote: Thu Apr 22, 2021 1:47 pmAlso would fitting a modern voltage regulator, e.g. a Murata, have any effect on this? Aside the being able to get rid of the heat sink that is.
Fitting a switching regulator / DC-DC converter reduces the current consumption from the mains PSU, which actually makes the problem worse!

The problem is the design of the DC-DC converter circuit formed by TR4, TR5, the coil and associated other components. TR5 compares the voltage on the +12V line (via resistors R58 and R59 which act as a voltage divider) to the +5V line (which here is being used as a reference voltage).

However, prior to the issue 4 boards, DC current could flow directly through the coil from the +9V (nominal) supply rail and feed the +12V line. This is not really a problem for the +12V rail if the +9V line is actually at, or near +12V (although any AC ripple on it from the unregulated PSU is undesirable).

However, if the 9V line is actually at 11.9V (approximately) or above, TR5 will start to turn off, and it will reduce the drive to, or even shut down TR4. TR4 drives pulses through the coil. With insufficient pulses from the coil, the coil will not be able to charge and discharge capacitor C46. Hence the system will not be able to generate the extremely critical -5V supply. Without this negative -5V supply to correctly bias the 4116 (or equivalent) DRAM chips, they are likely to be damaged.

If a switching regulator / DC-DC converter is fitted, the reduced current consumption from the unregulated UK1400 (or similar) PSU WILL result in the output voltage on the +9V (nominal) rail increasing compared to when a 7805 voltage regulator is used.

The solution is to modify the board with Sinclair’s latest revision of the DC-DC circuitry or to use a regulated +9V PSU. Although fitting this modification to an issue two board is considerably more difficult than in later issue boards.

If either the modification is done, or only regulated +9V PSUs are used, then, and only then, is it safe to change the 7805 voltage regulator to a switching regulator / DC-DC converter.

I hope this all makes sense...

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
Lardo Boffin
Posts: 2160
Joined: Sat Nov 26, 2016 2:42 am

Re: Poorly issue two Spectrum

Post by Lardo Boffin »

1024MAK wrote: Thu Apr 22, 2021 6:18 pm
If either the modification is done, or only regulated +9V PSUs are used, then, and only then, is it safe to change the 7805 voltage regulator to a switching regulator / DC-DC converter.

I hope this all makes sense...

Mark
Makes sense thanks Mark. Does this basic principal of using a switched mode supply apply to all issue twos that have not had the mod done?
ZX80
ZX81 iss 1 (bugged ROM, kludge fix, normal, rebuilt)
TS 1000 iss 3, ZXPand AY and +, ZX8-CCB, ZX-KDLX & ChromaSCART
Tatung 81 + Wespi
TS 1500 & 2000
Spectrum 16k (iss 1 s/n 862)
Spectrum 48ks plus a DIVMMC future and SPECTRA
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: Poorly issue two Spectrum

Post by 1024MAK »

Actually it affects issue one, issue two, and issue three boards, unless they have been modified during service or repair, or for the issue 3 onwards, they may have been modified during manufacture.

The issue 3B and later versions include capacitor C78 / C80 to prevent the DC current flowing from the +9V rail to the +12V rail. Although not all these PCBs actually have component positions for all the added components.

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
redgatemodels
Posts: 160
Joined: Tue Sep 05, 2023 11:56 am

Re: Poorly issue two Spectrum

Post by redgatemodels »

Following this one too as my S2 is poorly too ….
Post Reply