UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Closed Thread
 
Thread Tools
Old 22nd Jan 2023, 12:49 am   #21
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

Thanks for all your help. I'll have to study the circuits and try a few ideas. The 7447 looks a good starting point and the transistor output stages could be similar to those that drive Nixie tubes. I'll have a good look through my old circuit diagrams for some further ideas. It will take me a whille to fathom it out but definitely worth experimenting with.

Regards,
Symon
Philips210 is offline  
Old 22nd Jan 2023, 1:29 am   #22
factory
Octode
 
Join Date: Jul 2009
Location: Cheshire, UK.
Posts: 1,706
Default Re: Philips PM6614 Frequency Counter

One last thing, to match the font of the DM8884A, which has tails on the 6 & 9, the 74LS247/248 would match the font, again only the 247 is still made.

David
factory is online now  
Old 22nd Jan 2023, 1:34 pm   #23
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

That's a good point David, thanks for pointing it out, I had overlooked that.
I see what you mean, for instance the numeral '6' on the 7447 doesn't use the top 'a' bar but the top bar is used in the PM6614's display for the 6. It looks like I'll start off with the 74LS247.

Regards,
Symon
Philips210 is offline  
Old 23rd Jan 2023, 5:30 pm   #24
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

Thinking about typical multi-digit 7 segment displays, it's usual to have a decoder/driver IC for each display. In the case of this counter using the Panaplex display, a single decoder/driver IC drives all displays. Every digit's 'a' segment is connected to a common pin. It's the same for the other segments. Digit selection being controlled by applying HT to the relevant digit's anode.
Only one display lights at any instant due to the anodes being strobed, but all digits will appear to light at the same time due to the scanning speed. The timing has to be right to ensure the current BCD decoded data causes the right digit to light. This is a good example of a multiplexed display.

I can't say I've seen just a single IC driving multiplexed LED displays. Perhaps the brightness would suffer when being strobed.

The 74LS47 and 74LS247 ICs have inputs to suppress leading and trailing zeros. I think this applies when using several of these ICs in a typical set up. Not sure if that could be applied in a single IC application such as this.

I am going to build a breadboard circuit and have an adapter lead to plug into the DM8884A's socket. Current flow in the output driver transistors will have to be carefully checked to avoid damage to the display. Keeping the existing 2k2 resistors in circuit seems right. The blanking circuit could be a challenge to get right although it should work without any blanking as an initial test.

Regards,
Symon
Philips210 is offline  
Old 23rd Jan 2023, 6:09 pm   #25
Dickie
Octode
 
Join Date: Sep 2004
Location: St. Albans, Hertfordshire, UK.
Posts: 1,477
Default Re: Philips PM6614 Frequency Counter

Multiplexing 7-segment displays used to be common. Both the Racal 99xx series and Marconi 24xx series of counters used the method. Schematics of both are freely available online if you want to see how its done. Each display gets driven quite hard for a small time which means that if the multiplexing stops for any reason, the "lucky" digit is extremely bright!
__________________
Regards,

Richard, BVWS member
Dickie is online now  
Old 23rd Jan 2023, 6:31 pm   #26
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

Yes, it puts quite a strain on the lit display when the multiplexing fails.
Thanks for the mention of the Racal and Marconi circuits, I'll look those up on the BAMA resource.

Regards,
Symon
Philips210 is offline  
Old 6th Feb 2023, 6:23 pm   #27
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

Hello again.

I've been making some progress with a replacement circuit for the DM8884A BCD to 7-segment decoder/driver IC and have a working prototype.

Initially, I built a circuit using a 7447 IC. This has active low outputs which are open collector. In order to drive the high voltage npn output transistors for the Mullard Pandicon ZM1500-9 display, the outputs from the 7447 need to be inverted. I built a circuit on breadboard and it worked well but there was no blanking on adjacent digits.

I found the original blanking pulses that feed the DM8884A were not suitable to drive the blanking pin 4 of the 7447.

Looking at the block diagram for the DM8884A, it can be seen the blanking is applied in the output stages of the IC including blanking for the decimal point. I set about experimenting with NOR gates, taking the a to g segment outputs from the 7447 and applying the blanking signal to each gate. Again, I couldn't get the display to blank correctly.

Looking at the blanking waveform on the 'scope I found it to be incorrect. Its polarity was wrong and was very low amplitude so no hope of driving a TTL gate. A small modification in the PM6614's blanking circuit resolved this problem. TS139 was incorrectly biased for connection to a TTL circuit. The solution involved replacing R192 with a 100 Ohm resistor. A new 1k resistor is also added from R192 to ground. The circuit worked well. I revisited using pin 4 of the 7447 and this also worked well simplifying things somewhat. So no need therefore in adding blanking after the output stages of the 7447.
The decimal point also required blanking and that has been incorporated in my circuit.

To further simplify the circuit, I built one using a 7448 which has active high outputs and is thus suitable for directly driving the npn high voltage output transistors. This worked well but the display font is not the same as the original. I ordered a 74LS248 which resolved that.

One issue with the blanking was that although it correctly blanked the unwanted segments, the display was generally low brightness overall. I experimented with C140 in the differentiating circuit in the PM6614. In the original circuit, the blanking pulse is 12us wide. My assumption was that to reduce the blanking time and hence produce a brighter display, then the pulse width needs reducing. It was a wrong assumption as increasing the pulse width has had the desired effect. Increasing C140 from 820pF to 4n7F works well. A value of 3n3F is also acceptable with a marginally darker display.
I'm left wondering why increasing the blanking pulse width also increases the display brightness contrary to what I'd have thought.

Please see a few attached pics.

In the first, the working display but with no blanking. Note the partially lit segments.
The second pic is of the PM6614's segment blanking circuit with some waveform details.
The third pic outlines the effect of having a suitable collector load resistor to the blanking transistor TS139.
Pic 4 shows the circuit diagram of the prototype.
In pic 5 a correctly blanked display.

The next thing to do is make a circuit board to fit the available space in the counter.
So far it's been an enjoyable project. More is to follow in due course.

Regards,
Symon
Attached Thumbnails
Click image for larger version

Name:	no blanking.jpg
Views:	79
Size:	59.0 KB
ID:	272785   Click image for larger version

Name:	pm6614 blanking cct.jpg
Views:	82
Size:	32.4 KB
ID:	272786   Click image for larger version

Name:	pm6614 blanking cct mods.jpg
Views:	79
Size:	29.0 KB
ID:	272787   Click image for larger version

Name:	74ls248 cct.jpg
Views:	80
Size:	46.9 KB
ID:	272788   Click image for larger version

Name:	correctly blanked.jpg
Views:	77
Size:	68.7 KB
ID:	272789  

Philips210 is offline  
Old 23rd Mar 2023, 1:20 am   #28
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

Hello again

I've been making some progress with the replacement circuit for the DM8884A BCD to seven segment decoder IC. I built a circuit on matrix pad board using a 74LS248 as the decoder driver IC together with high voltage transistor output stages. Pics 1 and 2 show the completed board before installation.

From the outset, the challenge was how to mount this new circuit board in the PM6614. On the inside of the metal chassis there are some plastic slotted channels which lend themselves to the task. I made an offset bracket from 20 SWG aluminium which was secured to the matrix board with a couple of screws/nuts. The panel then had to be secured to the main board, another challenge. Originally there is a hexagonal spacer which extends from the main PCB to the inside of the top cover with a foam rubber buffer in between. It occured to me that this spacer could be replaced by two threaded spacers with my replacement circuit fitted between the spacers. Carefully measuring the original spacer and allowing for the thickness of the replacement board I was able to make the two spacers. No alterations have been made to the unit which helps to keep it original.

Pic 3 shows the new circuit board installed. I used solid core wire to make the connections to the DM8884A's IC socket, the individual wires which simply plug into the socket are quite secure. As previously mentioned, a couple of modifications to the original circuit are required namely R192, changed from 6k8 to 100R and also C140, changed from 820pF to 3n3. The components were removed from the main board and replaced with turned pin IC socket pins so that the revised value components can simply be plugged into the sockets. This allows the original value parts to be easily put back should a replacement DM8884A IC become available.

The other part of the repair involved the mains transformer. Again, I didn't want to make any new holes in the main PCB so made an alumininium plate with four mounting holes which correspond with the original tranformer's mounting hole locations. It is fitted to the main PCB with some nylon spacers in between.
I found a new toroidal mains transformer which looked a possible candidate but it would seem it is borderline regarding its VA rating. The transformer has two 9V secondaries each rated at 3.5VA. By connecting these in series gives an 18V 7VA transformer. Therefore the total allowable secondary current is 7/18 which equals 389mA. As the secondaries are feeding a bridge rectifier with capacitor reservoir, the maximum allowable DC current loading is 0.62 x 0.389 which is 241mA. I checked the DC current and this is around 250mA so very slightly over the limit. The other downside is the PM6614 has no ventilation holes. Ideally a transformer with a higher VA rating would be preferable. Initial tests so far show it is not getting hot but then the top cover hasn't been fitted. On the plus side it is mounted on a 14 SWG aluminium plate which will help a little. I'll have to do some more tests to prove whether it will be suitable.
Some P clips fitted to the mounting plate help to keep the wiring safe and tidy. I used a small tag strip for the secondary junction which is again fitted to the mounting plate. Pic 4 shows the transformer and mounting plate fitted to the main PCB. I hadn't fitted the crimp receptacle terminals to the red and orange secondary wires at this stage.
The original green class X mains filter capacitor which is fitted across the IEC mains inlet connector has also been replaced. Finding an axial lead class X2 cap isn't easy. I opted for a polypropylene 47nF 1000V DC rated cap which will be a safe replacement. Pic 5 shows an overall view.

The counter/timer is working well but I had an initial odd problem when testing. The PM661x series have a check facility by depressing the two Check buttons and setting the frequency range rotary switch, the display can be tested. I was finding a rogue digit 8 was appearing on each setting of the rotary switch. Also when fed with a sine wave at input A the frequency readings were 80% of their correct value. I wasn't looking forward to fixing this problem. After a fashion, I used the self check again and this time all was well so it was an intermittent fault. I had a little luck. I found as I applied slight pressure to the press button for channel B, I could envoke the fault. I sprayed the push button switches together with the rotary switch wafers with switch cleaner and worked the switches. This seems to have done the trick thankfully.

One other thing I've noticed. There's a high pitched whine which seems to be coming from the Pandicon display, perhaps linked to the multiplexing frequency. I can't really see how this can be solved. Perhaps it is normal for these counter/timers.

I have the top and bottom covers to respray as they look quite tatty. I also need to find a plastic foot to replace the missing one.

I'm quite pleased to get this old counter working. It's been a good project.

Regards,
Symon
Attached Thumbnails
Click image for larger version

Name:	IMGP8977.jpg
Views:	74
Size:	92.6 KB
ID:	275581   Click image for larger version

Name:	IMGP8976.jpg
Views:	70
Size:	88.7 KB
ID:	275582   Click image for larger version

Name:	IMGP8968.jpg
Views:	69
Size:	149.1 KB
ID:	275583   Click image for larger version

Name:	IMGP8969.jpg
Views:	69
Size:	113.5 KB
ID:	275584   Click image for larger version

Name:	IMGP8972.jpg
Views:	67
Size:	156.8 KB
ID:	275585  

Philips210 is offline  
Old 26th Mar 2023, 7:42 pm   #29
factory
Octode
 
Join Date: Jul 2009
Location: Cheshire, UK.
Posts: 1,706
Default Re: Philips PM6614 Frequency Counter

Well done at getting it going again, probably a good decision to use a through hole board, I've made some surface mount boards for a HP counter, the little SOT23 package transistors are the worst thing I've tried to solder by hand.

TI did a slightly different IC the SN75484 (DM8980) it has a very different pinout and has a latch function instead, but no comma function. The National databook does give more info of how they achieved the current programming function, one ref for the seven segments, a separate reference Zener for the decimal point and way too many transistors.
Click image for larger version

Name:	DS8980 IC from National Semiconductor Interface Databook 1979.jpg
Views:	31
Size:	40.6 KB
ID:	275800

David
factory is online now  
Old 27th Mar 2023, 12:23 am   #30
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

Hello David.

Thanks for your comments and info on the SN75484. I'll take a look at the data.

The odd thing with my counter is the noticeable whine that seems to be coming from the display. I'm wondering if one of the electrodes is loose within the display and it is ringing as a result of multiplexing/strobing. In my Series 1 counter, it uses the Mullard ZM1500-9 Pandicon display. Maybe these displays were prone to this rather than an actual fault.
The Series 2 counters employed the Burroughs Panaplex unit. I can't say if these displays also behave in the same way. Has anyone that owns one of these counters noticed any ringing in the display?

Regards,
Symon
Philips210 is offline  
Old 27th Mar 2023, 11:54 pm   #31
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

Hello again.

Please see a few attached pics.

Pic 1 shows the aluminium offset mounting bracket. Initially, I made a cardboard template to determine the height that the replacement circuit board should stand above the main PCB. The bracket was folded in a vice. It was easier to allow plenty to make folding easy. It was then trimmed to size.

Pic 2 shows the original spacer arrangement. It forms part of the fixing for the below board crystal oscillator sub panel.

Pic 3 shows the new spacer which is in two pieces to form the support and fixing for the new panel.

In pic 4, the old/faulty components. Many Philips green coloured caps often used in a flyback tuning role in TV receivers are prone to failure. The construction of the Class X types as mains filters are probably more reliable and safe but I replaced mine as a precaution with a 1000V dc rated polypropylene type.

The next job is to tidy the top and bottom covers. They'll need repainting.

More to follow in due course.

Regards,
Symon
Attached Thumbnails
Click image for larger version

Name:	IMGP8986.jpg
Views:	39
Size:	51.1 KB
ID:	275894   Click image for larger version

Name:	IMGP8995.jpg
Views:	36
Size:	47.6 KB
ID:	275895   Click image for larger version

Name:	IMGP8996.jpg
Views:	37
Size:	42.2 KB
ID:	275896   Click image for larger version

Name:	IMGP8994.jpg
Views:	39
Size:	60.9 KB
ID:	275897  
Philips210 is offline  
Old 29th Mar 2023, 4:24 pm   #32
Philips210
Nonode
 
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,562
Default Re: Philips PM6614 Frequency Counter

I've just had a mishap with the counter while trying to find the source of the noise around the display. I was applying gentle pressure around where it's soldered to the main PCB. I thought I could hear a slight variation in pitch. I picked up a cotton bud and continued probing around when I heard a slight crack. I looked at the display and the 2nd most significant digit was out. I thought oh no or words not to be repeated here. I honestly thought that I wrecked the display for which it will be extremely difficult to find a spare. I was quite annoyed with myself having spent a lot of time repairing it. I looked at the end of the cotton bud and I saw a minute sliver of solder. That was the cause of the short. I'm normally very strict about keeping a clean bench having been caught in the past. I guess the cotton bud must have picked it up.

I returned to it after lunch and didn't have that much hope of repairing it.
I first checked the base/emitter and base/collector junctions of the digit selection transistors. No problems could be found there. I then did some resistance checks from the outputs of the SN7442 BCD to decimal decoder IC to ground. I found pin 7 was slightly lower than the other output pins.

I thought a quick test for the faulty digit was to connect a working output to the 2nd digit's anode. I was relieved to see the digit light up proving the display was OK after all. I was about to condemn the SN7442 but thought again about the transistor output stages. When I properly checked the faulty digit's anode drive transistor, TS181 (BSS68), I found it has severe leak across the collector/emitter junction. It was about 39 Ohms.

Replacement is easier said than done due to very small through hole plated solder connections. No attempt was made to remove the transistor with its leads. Working it back and forth it was snapped off. Its three lead stumps were then carefully desoldered. Desoldering braid with a good hot iron removed the residual solder from the holes.
I didn't have a replacement BSS68 but a likely candidate was a BF421. The pin outs are different so it had to be oriented in a different way. With it replaced, I was very pleased to see the display was back to normal.
Had the display failed then the project would have been a complete waste of time. Attached is a pic of the missing digit.

The old counter lives again, thankfully.

Regards,
Symon
Attached Thumbnails
Click image for larger version

Name:	IMGP8998.jpg
Views:	42
Size:	84.0 KB
ID:	275997  

Last edited by Philips210; 29th Mar 2023 at 4:36 pm. Reason: typo
Philips210 is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 3:02 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.