Pick and Place Machine TVM802A / TVM802B

[qihe_kim]

I have the exact same hum! And it's driving me nuts! and yes it's transformer, i took the side cover off and touched the transformer with a wood stick and it was vibrating like crazy. Odd thing for me is it only just started happening after moving to a new workshop. We have tried power filters but did nothing. I was told it might be because here in Australia our mains are 240V not 220V and if you run it off a special UPS that puts out 220V it might stop it?
Could maybe remount the transformer on some rubber dampers but I think to get to it you need to pull the whole top desk off 

Thanks a lot for confirming this! Here in Germany we have 230-240V. A transformer with such an amount of hum is clearly in saturation. Dampening might help, but I guess it would be better to replace it. It is certainly not the 'positive pump' as stated by qihe_kim: My unit does not have one.

Hi ubbut,
Each machine from our company has the positive pump.And it is okay that machine is connected to the 240V power supply.
Could you send your e-mail by message?I can provide you process of checking the screw of the positive pump.
And it is best to send the Serial Number of you machine.

[AndyBig]

Hello everybody!

Please, can anyone help me? I purchased the TVM802BX, but only the program in Chinese is included (v 3.11). Can someone give the latest version of the software in English?
Thank you in advance.

[AndyBig]

The request is canceled, thanks to the manufacturer - sent me version 3.11 in English

[mgrossm1@yahoo.com]

My unit has a new problem. It is incorrectly placing all parts offset by a constant amount (1-2mm in X/Y). It is able to successfully detect the fiducials, pick up parts from the feeders and do optical recognition. I calibrated the nozzles. Can anyone think of a reason why this is happening? I installed a replacement mainboard and re-entered all of the settings. Is there a hidden menu I need to set?

[mgrossm1@yahoo.com]

I may have corrected the offset problem by adjusting the Nozzle 1 X/Y offset. It seems to affect part placement with or without vision. Very strange.

On a different topic, is there a good systematic way to adjust the feeders?

[ar__systems]

My unit has a new problem. It is incorrectly placing all parts offset by a constant amount (1-2mm in X/Y). It is able to successfully detect the fiducials, pick up parts from the feeders and do optical recognition. I calibrated the nozzles. Can anyone think of a reason why this is happening? I installed a replacement mainboard and re-entered all of the settings. Is there a hidden menu I need to set?

No, it is just the s/w is weird.

[ubbut]

Been using the machine now for a while and have some feature requests. Please consider implementing these, as this would greatly simplify working with the machine. Thank you (I have software version 3.09)

• Most importantly: instead of a global 'Down to stack' parameter, please introduce a different one for each stack. At least for the IC-stacks this is mandatory. Instead of having to build all trays to the same hight, just define it in software. Please
• It would be absolutely great if we could use camera 2 to calibrate the feeders. Part position and prick offset. This would be much faster and simpler then the current way of using the nozzle and do several test pricks
• When configuring a stack, only nozzle 1 can be used. This is not great, because I would like to be able to test pick-up with nozzle number 2. Please include this feature.
• At least in my software version (3.09) Move mode is restricted to 'nozzle' in the IC stack configuration, 'vision' can not be used. Please make this possible. Also again, we can only use nozzle 1 here, It would be great if nozzle 2 was selectable. if have had situations where I configured a stack with nozzle 1, only to find that when picking with nozzle 2, the machine will run into its end-switches
• Speaking of end-switches. If the machine knows where it is and how large it is - is it really necessary to run it all the way into the end switch. Could it not stop before? (And not create the laud beeping sound)
• At least in my particular machine there is a strange bug: If I use the manual controls to move down nozzle 2, when it reaches the limit it is not stopped by software. Rather (the belt) slips and both nozzle 1 and 2 position will be wrong. Only a complete shutdown of the machine fixes this

Thank you very much !!

[ubbut]

On a different topic, is there a good systematic way to adjust the feeders?

I would love to know that, too. I start defining the parts positions with the nozzle), then test the prick movement. However often after the prick movement, the part is in a different location and I have to define it again. This is quite confusing.

[AndyBig]

Hi!

instead of a global 'Down to stack' parameter, please introduce a different one for each stack. At least for the IC-stacks this is mandatory.

Parameter 'Down to ... ' for the IC-trays is 'Down To Front' (see in attachment), it can be configured separately from the feeders

It would be absolutely great if we could use camera 2 to calibrate the feeders.

But the camera2 can be used to configure feeders. For this, there is a "Vision" switch and a "Vision MoveTo" button (see attachment).
In general, setting up feeders is a fairly simple process when you begin to understand its principle. But it takes quite some time for all positions.

At least in my particular machine there is a strange bug: If I use the manual controls to move down nozzle 2, when it reaches the limit it is not stopped by software. Rather (the belt) slips and both nozzle 1 and 2 position will be wrong. Only a complete shutdown of the machine fixes this

This is a lack of hardware - there is only a sensor for the upper position of both nozzles, but there are no sensors for the lower position. However, it is unclear why this sensor does not work when the nozzle is raised after the nozzle's normal position has been disturbed.

My software version is 3.11, machine is TVM802BX.

[ubbut]

Parameter 'Down to ... ' for the IC-trays is 'Down To Front' (see in attachment), it can be configured separately from the feeders

Yes but not for each IC-Tray separately. If I have several trays, they all need to be the same hight...

But the camera2 can be used to configure feeders. For this, there is a "Vision" switch and a "Vision MoveTo" button (see attachment).
In general, setting up feeders is a fairly simple process when you begin to understand its principle. But it takes quite some time for all positions.

You are correct of course. I think I should have specified that it would be great to also adjust the Prick Correct with the camera.

[AndyBig]

If I have several trays, they all need to be the same hight...

But you can adjust the height of each component, which will be more correct, I think This height is reflected both on the installation on the PCB and on the capture from the tray.

I think I should have specified that it would be great to also adjust the Prick Correct with the camera.

Yes, here you are right, it would be more convenient to correct the prick with the camera too. Now we have to move the prick two or three times for a satisfactory correction of the prick. But it's better to check not the nozzle, but the camera

From myself, I can add something that I really miss when calibrating feeders - divisions on the camera lines for more accurate alignment of the crosshairs at the center of the component. Now I have pasted on the screen strips of transparent scotch with the marks drawn by the marker (in the attachment).

[ubbut]

If I have several trays, they all need to be the same hight...

But you can adjust the height of each component, which will be more correct, I think This height is reflected both on the installation on the PCB and on the capture from the tray.

This sounds more like a work-around. Also quite sure it does not work. Imagine I have a tray sitting 10mm lower than the other one. I define the component height to be 10mm/11mm. It will be picked up ok, but then dropped in mid-air over the PCB

[AndyBig]

Probably, it means that all trays of the same height I print trays on a 3D printer, one for all the necessary components. The bottom is at the same height, but components of different heights and this height I specify in the settings of the components.

[mgrossm1@yahoo.com]

Discovered the prick solenoid mounting screws were loose. Needless to say, realigning and tightening them has greatly improved the feeder reliability.

[mgrossm1@yahoo.com]

I just did a run of boards and it appears that parts are shifted toward the left with increasing Y coordinate (looking at the front of the machine). Does this imply the Y-Scaling value (currently set to 1) needs to be increased?

Thanks,
Mike

[AndyBig]

I just did a run of boards and it appears that parts are shifted toward the left with increasing Y coordinate (looking at the front of the machine). Does this imply the Y-Scaling value (currently set to 1) needs to be increased?

Have you put several separate boards? If so, then the reason may be in the not very precise external size of these boards. It is better to check on the ruler instead of several boards or on one big board.

[mgrossm1@yahoo.com]

No, just one board.

[AndyBig]

No, just one board.

Then yes, you need to increase the value of scaling, but in X, not Y, since the components are shifted to the left. I correctly understood - the more the component is located to the right, the more it is shifted to the left?

[mgrossm1@yahoo.com]

Yes, you are right. It is the x scaling.

Thanks,
Mike

[mgrossm1@yahoo.com]

After verifying the X/Y scaling is correct with a ruler, I ended up slightly modifying the fuducial coordinates to fix the placement issue. I don't like the solution very much but is works. Can anyone think of another adjustment to compensate for parts being shifted with increasing X value?

[FreezeSSC]

Hello everyone,

Was looking into getting my own PnP machine I do small runs (50 boards a year or less) of a heavily populated 11" by 7" PCB, just one sided mostly all 603, SOIC, TSSOP and SOT components.  I have about 4000 Components per board, would you guys recommend something like the TVM802B for my use case?  Looking for general feedback as this is new territory for me, I pay a lot for a local PnP shop so this could save me money in the long term but if its something that's constantly breaking or that has a lot of quirks I may just back off.

[ubbut]

Hello everyone,

Was looking into getting my own PnP machine I do small runs (50 boards a year or less) of a heavily populated 11" by 7" PCB, just one sided mostly all 806, SOIC, TSSOP and SOT components.  I have about 4000 Components per board, would you guys recommend something like the TVM802B for my use case?  Looking for general feedback as this is new territory for me, I pay a lot for a local PnP shop so this could save me money in the long term but if its something that's constantly breaking or that has a lot of quirks I may just back off.

How many different components / board ? Mostly 0603 / 0805? TVM802 runs mostly smooth when set-up correctly, although I like to watch in case something goes wrong. It's not a fast machine by industry standards, but if you can afford to wait a bit it is great. At 100% speed it surprisingly fast IMHO.

[FreezeSSC]

I have 76 different components per board.  The majority are smd 0603 with some 1206. 

[ubbut]

I have 76 different components per board.  The majority are smd 0603 with some 1206.

TVm802B has 48 feeders. Even if you manage to use them all and put some components in trays, you'll have to load the machine twice with such a board. Maybe you can get away by placing the less used components by hand. That depends on the number of boards of course. You'll have this problem with every machine and it is probably the reason your assembly house charges so much.
Usually it is possible to reduce the number of components: A pullup does not have to be 100k for example.

[mikeselectricstuff]

I do small runs (50 boards a year or less)

Hard to see how it would make sense doing that few in-house

have 76 different components per board.  The majority are smd 0603 with some 1206.

The first thing to look at is whether you can reduce that BOM. That would also save costs on subcontract assembly.