On 21 May 2012 at 11:14, David Griffith wrote: My gosh yes. If people can build piano-roll readers, a paper tape reader would be child's play. --Chuck

On 21 May 2012 at 14:36, Richard wrote: Well, there's still a guy selling reader kits: http://www.msdsite.com/OLA/TR01/tr01.html You do need a parallel interface. AFAIK, no Bluetooth model yet. --Chuck

I could do that with a leg on my back, but I never saw in person a perforated tape. Is it so hard to find/expensive to find? If it is cheap enough, I'd love to buy some tape and do a nice reader :)

You do not need a sproket wheel, it fact it's better (for the tape) if you don't use one. What you need is the photodiode array. Paper tape is friendly in that firstly there is a 'clock' for each dta word, namely the sprocket track. The holes i nthat are smalelr, so you can use them (with a smaller apperture sensor) as a clock for the data holes. The second friendly feature is that the holes are on a 0.1" pitch in both direcitons, prototpying board makes a usful jig for making parts. Now, the pphotidode array could be a problem. If you can find some photodiodes (or phototranssitors, slower, but that may not be a major problem) which are small enough, you cna mount them on 0.1" centres. if not, then you can use a trick that was done in an early Elliot reader which seems to use OCP71 phototransitors, namely fibre optics. The Elliot has glass rods and light guides from the read head to the phototransistors. These days you could use plasic fibre optic cable and suitable snesors. That's the hard part. The other part is moving the tape. This is hard if you want to be able to stop on a character, but as I said last night, for archiving tapes this may not be a problem. In the latter case, a simple capstan and pinch roller will feed the tape past the head, or you coudl even just turn the takeup spool (with a week brake o nthe feed spool) if you don't mind the data rate changing. -tony

If the person in question doesn't already have either one, I agree. But if you already have a scanner, known to work, and don't have the parts you mention, or don't have the "model engineer's workshop" tools...then I have trouble seeing anything wrong with using the scanner. That would, for example, be my own case - I probably could kludge something together to pull paper tape past phototransistors, but (a) it would be a kludge, (b) I don't have the phototransistors, and (c) I don't have (I would have to design and build - not difficult, but it would have to be done) the electronics to interface the thing to a computer. I do, however, have a scanner that needs nothing more than plugging in of wires to work fine. /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTML mouse at rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

* On Wed, May 23, 2012 at 07:18:24PM +0100, Tony Duell <ard at p850ug1.demon.co.uk> wrote: [...] I agree that home-made paper tape readers seem pretty easy, as projects go. But what I'd REALLY like to see is a home-made paper tape PUNCH. That's got to be tricky, even for a machinist! :) -Seth

It's a lot harder than a reader, but I don't think it would be imposible. I've worked on many commercial paper tape punches, and most, if not all, of them could be reporduced by a good model engineer (== home shop machinist). There seem to be 2 basic ways to make such a mechanism. The Facit 4070 typifies one, the GNT34 ( and BRPE) the other. The Facit 4070 has a rodary solenoid and link rod for each punch pin, and a steper motor/capstan to move the tape.The entire punchign force comes from the solenoid The latter haev a motor drive mechanism, and a set of small solenoids (in the GNT34, they llok like the coils from 'continental cradle relays') whci engange the mechansim with each pin, in many cases by sliding a spring-steel 'interposer' bekhind the pin, thus forcing the pin to be moved when the mecanism does. I am not sure which would be easier to make. However, the reader is probalby the more useful device in that it lets you archive paper tapes onto modern machiens nd thus to the web, etc. If anyone wants to run them on the oriignal hardware and doesn't have a punch to recreate paper tapes, it wouldn't be hard to make a paper tape reader emulator (reding the data from something like an SD card) for the classic computer. -tony

That sounds about right. Only if you're unimaginative enough to not loop the tape back and scan multiple segments of it per scan. I have neither scanner nor paper tape at ready hand, but I'd guess that I'd get some eight to ten lengths of tape side-by-side on the scanner bed. Of course, that's only a factor of some eight to ten improvement. Well, sure. We already knew the right answer for you was to build the reader. :-) I didn't mean that that made it impossible. It just is one more raising of the bar as compared to things I do have on hand. I considered mentioning that. I didn't, because I'd need to write some host software either way: either software to process the scanned images or software to interface to whatever hardware interface I build. Of course, for most people, one or the other of those would be easier, probably depending mostly on what software the person in question is used to writing. /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTML mouse at rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

If you don't have said tools, the solution is much the same as for me if I needed a scanner. There are basically 4 options : Build them Buy them Borrow them Get sometody else to do it. I think we can forget about 'build them'. It is possible to build a small lathe at home, people have done it. But mosto f hte time said people are (a) already knowledgabl abotu such tools and (b) have acecss to a lathe, etc ot make soem of th parts on. So : 'Buy them'. The only major expense would eb a lathe with milling facilities. The parts of a paper tape reader as amall enough that you could make them on soemthing like a Unimat. That would cost about as much as a good PC + scanner +... setup. And I suspect it would be useful for a lot longer. A larger lathe is going to cost as much as a small car, but it'll easily outlive said car :-) 'Borrow them'. Not as silly as it sounds. It looks like some hackerspaces [1] have access to small machine tools. AS do some model engieering clubs. I would think a paper tape reader would eb a suitlabe project for such organisations. 'Get soembody else to do it'. Maybe you have a friend who makes model steam enginers, etc but can't get his PC working, or needs some bit of electronics fixed, or needs advice on wiring, or... Offer to do that job for him if he'll make the capstan for your reader on his lathe. [1] I am not a member of any such organsisation, so I am going on what their websites say. Perhaps I should join the local-ish one, but I don't really see what I would get out of it,given that i have most of the equipment theu offer in my home workshop. Does anyone have any thoughts? -tony

Okay, folks are suggesting phototransistor+light source for this kind of thing. But I was wondering... Does anyone know of a tape (or card) reader that uses a slight vacuum (or positive pressure) (player piano style) to read the patterns? Such a method wouldn't be sensitive to the material (e.g. clear mylar) used for the medium. Not that I'm planning to build one, but that I'm curious. --Chuck

On Wed, 23 May 2012, Chuck Guzis wrote: Fast paper tape readers (like the Facit PE1000) have a capacitive read head. It generates a HF, and the paper tape acts as dielectricum that changes its constant depending on hole/no hole. This change is then detected by the read circuitry. Christian

On Thu, 24 May 2012, Tony Duell wrote: Yes, I have the successor to the PE1000, relabeled by Telefunken as LSL100 (for the TR440 system). I had scanned the service manual for the earlier model (no schematics there, but I've drawn them by hand, but I need to find and scan them); it's on bitsavers. More documents for the PE1000 can be found here (from Philipp): http://bitsaving.hachti.de/box1/facit/ Especially the "Reader Manual" contains functional description and schematics. Our model does 500 or 1000 cps. But apparently there existed some models reading up to 2000 cps (e.g. Telefunken LSL200) or even more. But then you really have a problem with tape feeding and winding. (At 1000 cps the tape just shoots at least 1m horizontally out of the reader if you don't have a fast and accurate motor winder). Christian

I've nto seen one, but that doesn't mean it was never tried. In gneral it's easier to read paper tape with a vacuum in that it keeps the tape against the read head... The common metods used to read paper taper were : Mechncial ; I've not seen one that actually makes contact through the holes, but mechancial pins (often called 'peckers' over here) that went throug hthe holes and operated contcts if said hole was there wewre common. Normally the peckers were in a line perpendicualr to tape movement and the tape was moved in steps, but at least one Creed serial tape reader moved the tape continuously nd lifted the peckers in sequence to read ech data trcak in turn, the mechansim then operated a single contact depending on the state of the crurent pecker, thus giving a serial output Optical : We all know how that works. And that it doesn't work for transparent tape. Trend hand a nice design that had 2 sensorts on the sprocket track, 2.5 characters apart. If the one in-line wit hthe data was brighter than the other one by a certain threshold, that meant there was a sprocket hole at the read head, the (weighted?) mean of the 2 sprocket-track sensorts set the thresholds for the data track sensors. It was pretty much self-adjusting. Capacitive : I've never seen one, but apparently Facit tried a thing that detected the cahnge in capacitance when the tape (rather than a hole) was over the read electrode. I am told it was not that reliable... -tony

That is a lot faster than laying the tape across a flat-bed scanner.

Groan! I haev no intnetion of trying to measure the electrical conductivity (in siemens (==mhos) as I susepct I would not find the experience very pleasant to say the least. -tony

Just for the lulz, yesterday I saw a current amplifier that works in the order of FEMTOamps (!!!) and uses - gasp - 10 GIGAohms resistors :oO Things I never saw in my life (used in gas chromatographs) World and life are full of interesting things :D

Wow! I only believed it because I saw with my own eyes. The circuit is built on top of teflon insulators, the fiberglass board has no enough insulation for this kind of amp. I wasn't able to take photos (is a trade secret of one of my clients) but I was very impressed nonethless for the building quality and the values of parts. It is very funny to see a resistor encased on a glass tube :oD These are things I never imagined seeing :oD

I think the larged 'normal looking' resistor with coloured bands that I've seen/used was 22M, Maybe 56N I've certainly used 1G resistors, but for measuring instruemtns, not in classic computers (or consumer electronics). Thetes' one place you might well find high-value resistors, and that's in the EHT circuits for CRTs. An EHT voltmeer can be nothing more than a microammeter and a high-value resistor in series, in fact the Heathkit one that I have (the only Heathkit I ever built...) is just a 50uA moving-colil meter and an 800M resistor (it measures up to 40kV, of course). That reissotr is a glass tube about 6" long, in the barrel of the probe. It has femae threads at each end, the probe spike screws into one of them, the other goes onto an adapter that's wired ot the meter. Some of the better monitors (like my Barco and the one on my HP9836CU) have regulated EHT supplies. THis typically involves a portential divider conencted across the (25kV) supply to the CRT anode, with the tap (at a nice low voltae, around 12V, say) goign to the error amplifier that controles the EHT supply. Since the currents inolved are necessrily low, the 'top' reisistor is a high-value thing. It also has to be well-insulated since it's dropping nearly 25kV. It doesn't look like a reissotr, it's a potted modlue with the EHT cable going in one side and a normnal wire coming out the other going to the rest of the cirucity.. My scheamtic for the HP monitor I mentioend says I measured it (it wasn't marked) as 200M. -tony

Resistors get much, much larger that 10 megaOhms. Much larger than 10 gigaOhms, even. When they get that large in value, they need special handling, as the oil from fingers can throw the value way off. I live near, and have been to, Hipotronics, where they fool around with resistors in the teraOhm and capacitors in the megaVolt range. -- Will

Great article in link. When you get down in the dirt, all the way down, It takes good layout engineering to keep noise down and the signal over the thermal and coupling noise. I didn't know super insulators accumulated that much static/free electrons, but it makes sense. It's more black art then RF design. Must read. Jim Dave McGuire wrote:

On Fri, 25 May 2012, Dave McGuire wrote: ??? You have to translate that for me as I don't understand what you mean... Christian

Hmm... I wonder what the advantage is over optical units, then... I have some Trend optical readers. The HSR500 is a 500 cps bidirectional unit, the UDR700 is a Unidirectional 700cps unit (there was also a slower UDR350). These readers will get up to full speed in quite a short time (the motor runs continuously, the tape motion is started by a soleoid lifitng a pinch roller onto the tape), and when running at full speed you can stop them _on_ a chracter (in other eords, you can read a character, and if your controller is fast enough to determien that's the character you want, and then deassert the tape run signal, the tape will stop without any more characters being read, if you then assert the tape run line again, the next sproket pulse on the interface conenctor will be the next character on the tape, nothing is lost). These readers are very simple, both electroncially and mechanically. I wonder why cpacitive reading is a good thing. -tony

Err, yes... And not a description of a bird that is feeding, right :-) We have an experession here 'Keep your pecker up' which means essentially 'Stay cheerful'. That has as a rather different meaning to you, I guess :-) But the term 'pecker' in relation to mechancial tape readers is used in official Creed manuals and books about such machines. Mind you, I have at least one paper tape punch manual that describes a swithc that feeds totally unpunched tape (not even sprocktt holes) and producing 'virgin tape -- that is tape without holes'. That could have other connotations :-) -tony

Would you rather call it a "wanker"?

I did suggest a Unimat or similar. That's a fairly small machine (it goes o na normal bench) and can be packed away when not in use. That makes it a lot less capable and less accurate than a larger lathe, but from what I've seen it's a lot better than having nothign at all. Yes, we'd all like a large lathe and a milling machine, a 128 channel 1GHz+ logic analyser, 8 channel 1Gsample/sec DSO, 10 digit DMM and so on. But you have to make compromises and having useable tools and instruments is a lot better than nothing. -tony

Probably not... Being able to resolve 1uV in 10kV would be pretty amazing :-). But that doens't mean I can't want one :-) Actually an 8.5 digit DMM would be nice (or even a 6.5 digit one). -tony

Indeed :-) I acutally considered buying one at one point (not that I could really afford it, but...). What put me off was, as ever, the lack of a real service manual. I I bought soemthign that expensive and of that quality, I'd want to be able to keep it running. -tony

[HP/Agilent 8.5 digit DMM] quality, > > I'd want to be able to keep it running. I am sure they're very reliable, but wil lit still be workin without problems in 30 years time? It's not a cheap instruemnt, and I could see it being usful long into the future if it could be kept working. Of ocurse there is the slight problem of 'what the heck do you use to calibrate it' [1] Alas I have to. I can't afford to buy such an instruent unless it's going to last me a very long time. [1] Actually, I think I know one answer. A Josephon Junction. If you have such a drive in a microwave field (a dew GHz) and vary the curent through it, the votlage across it changes in discrete steps. The voltage change on eeach step depends only on fundamental contants nad the microwave frequency, so there's realy only one thing you haev to be able to measure. And frequency is one thing you can measure quite accurately On the otehr hand, getting a feqeuncy standard that's got to that sort of accuracy is not trivial at home. Nor is getting a liquid helium cryostat to put the junction in. -tony

Is this a secret or we may know that? ;oD

Hi It is no secret. I've posted it to this group before. First, consider how useless an ohm meter is. Usually something like a shorted capacitor is maybe 20 times the resistance of the ground plane. Looking for differences in the resistance across the ground plane is usually a waste of time. Most just use the high current method of making the shorted part hot or smoke. This is not always successful and can damage the board. The trick I use puts no current through the short. There is no need. First lets describe the problem if the short is between two traces and then the more complicated case that it is between to planes. Take a bench supply that can produce a couple amperes in the current regulated mode. Pick one of the two trace and connect the supply from end to end on that one trace ( we'll call it trace1 to keep from getting confused ). Place one fixed lead of the voltmeter on trace2( location isn't important ). With the other lead, of the meter, probe along trace1. When you reach the location where the meter reads zero volts, you've found the short. In some cases, there may be a lot of shunting resistance In this case, gradually move the contact points that you connect the power supply towards the zero volt point and also move the fixed meter lead. As you move the supply leads in, you'll pass the shunting resistances and get a truer location of the short. It is rarely necessary to do this on most solid state circuits. Now for the slightly more complicated case of power planes. Place the supply leads at two diagonal corners of the board. Use more current on a power plane( 5-10 amperes is usually good ). With the current across plane1, place one fixed lead on plane2. Now probe along Plane1 until you find the zero volt location. In this case, it usually forms a curved line. I usually just lay pieces of tape along this zero volt line. The short is someplace along the line. Now, move the power supply to the other two corners of the boards plane1. Repeat the finding of the zero volt line. Where this line and the previous one cross is the short. I've used this method in a production environment and maybe found several hundred shorts. It does take a sensitive meter, that is why I go for a 5 digit. I suppose it would work with a galvanometer used on a bridge as well but never tried that. At one time, I'd considered making a high current sine/cosine current source. One could then use an oscilloscope and just probe for a minimum circle or elipse. This would speed up the power plane locating. Dwight

On Mon, May 21, 2012 at 11:14:06AM -0700, David Griffith wrote: And it would need a few orders of magnitude less processing power as well ;-) Yes, with modern machines and for many tasks, machine cycles are cheap - but go for brute force when there is a better way? Kind regards, Alex. -- "Opportunity is missed by most people because it is dressed in overalls and looks like work." -- Thomas A. Edison

get a tripod and a camera and take photos then right some program that process's through them all and blam u got a way to archive ur tapes On Mon, May 21, 2012 at 1:21 PM, Dave McGuire <mcguire at neurotica.com> wrote:

Hi I have a high speed reader to but I've had troubleswith rubber bands. They leave gunk on the tapes andlayers stick together. On a high speed reader, itrips the tape before one has a chance to stop it. Mine is a parallel one and I just connected it to theprinter port ( guess that dates my laptop ). A simpleprogram reads the data and puts it into files. On making a reader, the ideas of using a pile ofphotodiodes to read a punch card doesn't makemuch sense. It might be better to have a LED lightat each hole position of a column and a single photodiode.One could then just sequntially drive the lightsand detect it with just one photodiode.One might need to aim each light at the diode.As with Dave, if someone has fan fold paper tape toarchive, I could do an reasonable amount. I'min San Jose, Ca, USA.I'm not setup to handle spooled tape safely.Dwight

I have a couple of commercial ahnd-cranked rewinders, and a couple of home-made (not by me) ones. The latter seem to have been made from hand-cranked gridners (the sort of thing used to sharpen knives) with the grinding wheel removed and a suitable disk.hub added. My Facit 4070 punches have a powered take-up spool. It was a simple modification to conenct the Ext switch (which normally simply grounds a pin on the interface conenctor and is raely used for anything) to winder motro so that pressign said switch caused the takeup spool to spin. Then just use that to wind up the tape. -tony

On 22/05/2012 19:44, Tony Duell wrote: Mine seems to be clever enough to stop without damaging the feed holes....

THis sounds like the read head for typesetter tape (which had offset sprocket holes) and amy not be much use for reading computer/telegraph tape. -tony